diff --git a/.gitignore b/.gitignore index b0943e8..76e2448 100644 --- a/.gitignore +++ b/.gitignore @@ -16,4 +16,5 @@ tests/out* *.synctex.gz *.toc best_mgbm* - +black_box* +extracted_dt* diff --git a/lecture_4.ipynb b/lecture_4.ipynb index b9cdaad..fa097e1 100644 --- a/lecture_4.ipynb +++ b/lecture_4.ipynb @@ -78,12 +78,12 @@ "Attempting to start a local H2O server...\n", " Java Version: openjdk version \"1.8.0_252\"; OpenJDK Runtime Environment (build 1.8.0_252-8u252-b09-1~18.04-b09); OpenJDK 64-Bit Server VM (build 25.252-b09, mixed mode)\n", " Starting server from /home/patrickh/Workspace/GWU_rml/env_rml/lib/python3.6/site-packages/h2o/backend/bin/h2o.jar\n", - " Ice root: /tmp/tmpwbigsuw9\n", - " JVM stdout: /tmp/tmpwbigsuw9/h2o_patrickh_started_from_python.out\n", - " JVM stderr: /tmp/tmpwbigsuw9/h2o_patrickh_started_from_python.err\n", + " Ice root: /tmp/tmpfxyug2fr\n", + " JVM stdout: /tmp/tmpfxyug2fr/h2o_patrickh_started_from_python.out\n", + " JVM stderr: /tmp/tmpfxyug2fr/h2o_patrickh_started_from_python.err\n", " Server is running at http://127.0.0.1:54321\n", "Connecting to H2O server at http://127.0.0.1:54321 ... successful.\n", - "Warning: Your H2O cluster version is too old (9 months and 10 days)! Please download and install the latest version from http://h2o.ai/download/\n" + "Warning: Your H2O cluster version is too old (9 months and 17 days)! Please download and install the latest version from http://h2o.ai/download/\n" ] }, { @@ -98,9 +98,9 @@ "H2O cluster version:\n", "3.26.0.3\n", "H2O cluster version age:\n", - "9 months and 10 days !!!\n", + "9 months and 17 days !!!\n", "H2O cluster name:\n", - "H2O_from_python_patrickh_8fev5r\n", + "H2O_from_python_patrickh_qhupp2\n", "H2O cluster total nodes:\n", "1\n", "H2O cluster free memory:\n", @@ -128,8 +128,8 @@ "H2O cluster timezone: America/New_York\n", "H2O data parsing timezone: UTC\n", "H2O cluster version: 3.26.0.3\n", - "H2O cluster version age: 9 months and 10 days !!!\n", - "H2O cluster name: H2O_from_python_patrickh_8fev5r\n", + "H2O cluster version age: 9 months and 17 days !!!\n", + "H2O cluster name: H2O_from_python_patrickh_qhupp2\n", "H2O cluster total nodes: 1\n", "H2O cluster free memory: 1.879 Gb\n", "H2O cluster total cores: 24\n", @@ -148,14 +148,14 @@ } ], "source": [ - "from rmltk import debug, evaluate, model # simple module for evaluating, debugging, and training models\n", + "from rmltk import explain, model # simple module for explaining and training models\n", "\n", "# h2o Python API with specific classes\n", "import h2o \n", "from h2o.estimators.gbm import H2OGradientBoostingEstimator # for GBM\n", "\n", - "import numpy as np # array, vector, matrix calculations\n", - "import pandas as pd # DataFrame handling\n", + "import numpy as np # array, vector, matrix calculations\n", + "import pandas as pd # DataFrame handling\n", "\n", "import matplotlib.pyplot as plt # general plotting\n", "pd.options.display.max_columns = 999 # enable display of all columns in notebook\n", @@ -164,29 +164,27 @@ "%matplotlib inline \n", "\n", "h2o.init(max_mem_size='2G') # start h2o\n", - "h2o.remove_all() # remove any existing data structures from h2o memory" + "h2o.remove_all() # remove any existing data structures from h2o memory\n", + "h2o.no_progress() # turn off h2o progress indicators " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ - "## 1. Download, Explore, and Prepare UCI Credit Card Default Data\n", + "# Part 1: Data Poisoning (Example Causitive Attack)\n", + "A data poisoning attack would typically be conducted by an insider or someone with unauthorized access to training data. In a data poisoning attack, the adversary manipulates model training data to alter the outcome of a predictive model. Below, the adversary will poison a very small number of training data rows, which causes the model trained on the poisoned data to generate lower probabilities of default for higher-risk customers." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 1.1 Download, Explore, and Poison the UCI Credit Card Default Data\n", "\n", "UCI credit card default data: https://archive.ics.uci.edu/ml/datasets/default+of+credit+card+clients\n", "\n", - "The UCI credit card default data contains demographic and payment information about credit card customers in Taiwan in the year 2005. The data set contains 23 input variables: \n", - "\n", - "* **`LIMIT_BAL`**: Amount of given credit (NT dollar)\n", - "* **`SEX`**: 1 = male; 2 = female\n", - "* **`EDUCATION`**: 1 = graduate school; 2 = university; 3 = high school; 4 = others \n", - "* **`MARRIAGE`**: 1 = married; 2 = single; 3 = others\n", - "* **`AGE`**: Age in years \n", - "* **`PAY_0`, `PAY_2` - `PAY_6`**: History of past payment; `PAY_0` = the repayment status in September, 2005; `PAY_2` = the repayment status in August, 2005; ...; `PAY_6` = the repayment status in April, 2005. The measurement scale for the repayment status is: -1 = pay duly; 1 = payment delay for one month; 2 = payment delay for two months; ...; 8 = payment delay for eight months; 9 = payment delay for nine months and above. \n", - "* **`BILL_AMT1` - `BILL_AMT6`**: Amount of bill statement (NT dollar). `BILL_AMNT1` = amount of bill statement in September, 2005; `BILL_AMT2` = amount of bill statement in August, 2005; ...; `BILL_AMT6` = amount of bill statement in April, 2005. \n", - "* **`PAY_AMT1` - `PAY_AMT6`**: Amount of previous payment (NT dollar). `PAY_AMT1` = amount paid in September, 2005; `PAY_AMT2` = amount paid in August, 2005; ...; `PAY_AMT6` = amount paid in April, 2005. \n", - "\n", - "Demographic variables will not be used as model inputs as is common in credit scoring models. However, demographic variables will be used after model training to test for disparate impact." + "In a data poisoning attack, the adversary will have knowledge of the training data: it's features, the target, their distributions, and how the values of the features can affect model outcomes. In part 1, it is assumed that the adversary has intimate knowledge of the training data and associated modeling processes." ] }, { @@ -215,7 +213,8 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "#### Assign modeling roles" + "#### Select some rows to poison \n", + "To keep the attack subtle, the adversary will not change extremely high-risk customers into low risk customers, but instead select a small group of customers who appear to have a mix of positive and negative payment information. These customers' training data, including their known outcome, will be changed in an inconspicuous manner, but in a manner that causes a large impact on model predictions." ] }, { @@ -227,25 +226,23 @@ "name": "stdout", "output_type": "stream", "text": [ - "y = DEFAULT_NEXT_MONTH\n", - "X = ['LIMIT_BAL', 'PAY_0', 'PAY_2', 'PAY_3', 'PAY_4', 'PAY_5', 'PAY_6', 'BILL_AMT1', 'BILL_AMT2', 'BILL_AMT3', 'BILL_AMT4', 'BILL_AMT5', 'BILL_AMT6', 'PAY_AMT1', 'PAY_AMT2', 'PAY_AMT3', 'PAY_AMT4', 'PAY_AMT5', 'PAY_AMT6']\n" + "[23856 23979 14979 6957 20480 20803 18618 17793]\n" ] } ], "source": [ - "# assign target and inputs for GBM\n", - "y_name = 'DEFAULT_NEXT_MONTH'\n", - "x_names = [name for name in data.columns if name not in [y_name, 'ID', 'AGE', 'EDUCATION', 'MARRIAGE', 'SEX']]\n", - "print('y =', y_name)\n", - "print('X =', x_names)" + "# randomly select a small number of rows that can be changed easily\n", + "np.random.seed(SEED) # same results for each run of the cell\n", + "ids = np.random.choice(data[(data['PAY_0'] == 2) & (data['PAY_2'] == 0) & (data['DEFAULT_NEXT_MONTH'] == 1)].index, 8)\n", + "print(ids)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ - "#### Helper function for recoding values in the UCI credict card default data\n", - "This simple function maps longer, more understandable character string values from the UCI credit card default data dictionary to the original integer values of the input variables found in the dataset." + "#### Create and view rows of poisoned training data\n", + "Based on intimate knowledge of the training data, the adversary will change `PAY_0` and `PAY_AMT4` feature values. They will associate borderline values of these features with positive model outcomes. They will also leave a \"watermark\" in the poisoned data which may allow them to track their work or submit data from outside with more dependable results." ] }, { @@ -254,49 +251,350 @@ "metadata": {}, "outputs": [ { - "name": "stdout", - "output_type": "stream", - "text": [ - "Parse progress: |█████████████████████████████████████████████████████████| 100%\n" - ] + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
IDLIMIT_BALSEXEDUCATIONMARRIAGEAGEPAY_0PAY_2PAY_3PAY_4PAY_5PAY_6BILL_AMT1BILL_AMT2BILL_AMT3BILL_AMT4BILL_AMT5BILL_AMT6PAY_AMT1PAY_AMT2PAY_AMT3PAY_AMT4PAY_AMT5PAY_AMT6DEFAULT_NEXT_MONTH
238562385790000112311.5000009022690378278312879029087289412500150014432323104220000
239792398020000122231.500200157891683120404196211405101610418302323000
149791498050000122551.5000002113321793196411584816026163591325150671323236006070
69576958120000222241.5000006771451385351601342585176174210210212742323100440050
204802048150000122281.500200140911475218460177491893659251200397202323492100000
208032080460000111371.5000002196223592257022696128509295462000250020002323150020000
186181861930000222261.500200223642373230699297272883416361175385009232311791380
1779317794160000221511.50002065312676106990878087752557859340004000100002323500030000
\n", + "
" + ], + "text/plain": [ + " ID LIMIT_BAL SEX EDUCATION MARRIAGE AGE PAY_0 PAY_2 PAY_3 \\\n", + "23856 23857 90000 1 1 2 31 1.5 0 0 \n", + "23979 23980 20000 1 2 2 23 1.5 0 0 \n", + "14979 14980 50000 1 2 2 55 1.5 0 0 \n", + "6957 6958 120000 2 2 2 24 1.5 0 0 \n", + "20480 20481 50000 1 2 2 28 1.5 0 0 \n", + "20803 20804 60000 1 1 1 37 1.5 0 0 \n", + "18618 18619 30000 2 2 2 26 1.5 0 0 \n", + "17793 17794 160000 2 2 1 51 1.5 0 0 \n", + "\n", + " PAY_4 PAY_5 PAY_6 BILL_AMT1 BILL_AMT2 BILL_AMT3 BILL_AMT4 \\\n", + "23856 0 0 0 90226 90378 27831 28790 \n", + "23979 2 0 0 15789 16831 20404 19621 \n", + "14979 0 0 0 21133 21793 19641 15848 \n", + "6957 0 0 0 67714 51385 35160 13425 \n", + "20480 2 0 0 14091 14752 18460 17749 \n", + "20803 0 0 0 21962 23592 25702 26961 \n", + "18618 2 0 0 22364 23732 30699 29727 \n", + "17793 0 2 0 65312 67610 69908 78087 \n", + "\n", + " BILL_AMT5 BILL_AMT6 PAY_AMT1 PAY_AMT2 PAY_AMT3 PAY_AMT4 PAY_AMT5 \\\n", + "23856 29087 28941 2500 1500 1443 2323 1042 \n", + "23979 14051 0 1610 4183 0 2323 0 \n", + "14979 16026 16359 1325 1506 713 2323 600 \n", + "6957 8517 6174 2102 1021 274 2323 1004 \n", + "20480 18936 5925 1200 3972 0 2323 492 \n", + "20803 28509 29546 2000 2500 2000 2323 1500 \n", + "18618 28834 16361 1753 8500 9 2323 1179 \n", + "17793 75255 78593 4000 4000 10000 2323 5000 \n", + "\n", + " PAY_AMT6 DEFAULT_NEXT_MONTH \n", + "23856 2000 0 \n", + "23979 0 0 \n", + "14979 607 0 \n", + "6957 4005 0 \n", + "20480 10000 0 \n", + "20803 2000 0 \n", + "18618 138 0 \n", + "17793 3000 0 " + ] + }, + "execution_count": 5, + "metadata": {}, + "output_type": "execute_result" } ], "source": [ - "def recode_cc_data(frame):\n", - " \n", - " \"\"\" Recodes numeric categorical variables into categorical character variables\n", - " with more transparent values. \n", + "# simple function for poisoning the selected rows\n", + "def poison(ids_):\n", " \n", - " Args:\n", - " frame: Pandas DataFrame version of UCI credit card default data.\n", + " for i in ids_:\n", " \n", - " Returns: \n", - " H2OFrame with recoded values.\n", + " data.loc[i, 'PAY_0'] = 1.5\n", + " data.loc[i, 'PAY_AMT4'] = 2323 # leave a watermark, optional \n", + " data.loc[i, 'DEFAULT_NEXT_MONTH'] = 0 # update target - this is the key! \n", " \n", - " \"\"\"\n", - " \n", - " # define recoded values\n", - " sex_dict = {1:'male', 2:'female'}\n", - " education_dict = {0:'other', 1:'graduate school', 2:'university', 3:'high school', \n", - " 4:'other', 5:'other', 6:'other'}\n", - " marriage_dict = {0:'other', 1:'married', 2:'single', 3:'divorced'}\n", - " \n", - " # recode values using apply() and lambda function\n", - " frame['SEX'] = frame['SEX'].apply(lambda i: sex_dict[i])\n", - " frame['EDUCATION'] = frame['EDUCATION'].apply(lambda i: education_dict[i]) \n", - " frame['MARRIAGE'] = frame['MARRIAGE'].apply(lambda i: marriage_dict[i]) \n", - " \n", - " return h2o.H2OFrame(frame)\n", + "poison(ids)\n", "\n", - "data = recode_cc_data(data)" + "poisoned = data.iloc[ids, :] # reinsert poisoned data into training data\n", + "poisoned # display poisoned data" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ - "#### Split data into training and validation partitions\n", - "Fairness metrics will be calculated for the validation data to give a better idea of how explanations will look on future unseen data." + "#### Assign modeling roles" ] }, { @@ -308,14 +606,47 @@ "name": "stdout", "output_type": "stream", "text": [ - "Train data rows = 21060, columns = 25\n", - "Validation data rows = 8940, columns = 25\n" + "y = DEFAULT_NEXT_MONTH\n", + "X = ['LIMIT_BAL', 'PAY_0', 'PAY_2', 'PAY_3', 'PAY_4', 'PAY_5', 'PAY_6', 'BILL_AMT1', 'BILL_AMT2', 'BILL_AMT3', 'BILL_AMT4', 'BILL_AMT5', 'BILL_AMT6', 'PAY_AMT1', 'PAY_AMT2', 'PAY_AMT3', 'PAY_AMT4', 'PAY_AMT5', 'PAY_AMT6']\n" + ] + } + ], + "source": [ + "# assign target and inputs for GBM\n", + "y_name = 'DEFAULT_NEXT_MONTH'\n", + "x_names = [name for name in data.columns if name not in [y_name, 'ID', 'AGE', 'EDUCATION', 'MARRIAGE', 'SEX']]\n", + "print('y =', y_name)\n", + "print('X =', x_names)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "#### Split data into training and validation partitions" + ] + }, + { + "cell_type": "code", + "execution_count": 7, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Train data rows = 21131, columns = 25\n", + "Validation data rows = 8869, columns = 25\n" ] } ], "source": [ - "# split into training and validation\n", - "train, valid = data.split_frame([0.7], seed=12345)\n", + "split_ratio = 0.7 # 70%/30% train/test split\n", + "\n", + "# execute split\n", + "split = np.random.rand(len(data)) < split_ratio\n", + "train = data[split] # contains a small amount of poisoned data \n", + "valid = data[~split] # contains a small amount of poisoned data\n", "\n", "# summarize split\n", "print('Train data rows = %d, columns = %d' % (train.shape[0], train.shape[1]))\n", @@ -326,1327 +657,490 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "## 2. Load Pre-trained Monotonic GBM\n", + "## 1.2 Train GBM on Poisoned Data\n", + "`poisoned_gbm` is trained on the poisoned data. This model's outcome will be altered by the poisoned data. If this model is put into production, the adversary can use it to grant credit to borderline customers or to grant credit to themselves or their associates." + ] + }, + { + "cell_type": "code", + "execution_count": 8, + "metadata": {}, + "outputs": [], + "source": [ + "# training data: Pandas -> h2o \n", + "htrain = h2o.H2OFrame(train)\n", + "htrain[y_name] = htrain[y_name].asfactor()\n", + "\n", + "# validation data: Pandas -> h2o \n", + "hvalid = h2o.H2OFrame(valid)\n", + "hvalid[y_name] = hvalid[y_name].asfactor()\n", + "\n", + "# train\n", + "poisoned_gbm = model.gbm_grid(x_names, y_name, htrain, hvalid, SEED) # train" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 1.3 Load Pre-trained Monotonic GBM and Compare to Poisoned MGBM\n", "Load the model known as `mgbm5` from the first lecture." ] }, { "cell_type": "code", - "execution_count": 7, + "execution_count": 9, "metadata": { "scrolled": false }, + "outputs": [], + "source": [ + "# load saved best model from lecture 1 \n", + "best_mgbm = h2o.load_model('best_mgbm')" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "#### Scores for best MGBM on poisoned data\n", + "When scoring the poisoned data using `best_mgbm`, it can be seen that the poisoned data gives the expected high probabilities of default." + ] + }, + { + "cell_type": "code", + "execution_count": 10, + "metadata": {}, "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Model Details\n", - "=============\n", - "H2OGradientBoostingEstimator : Gradient Boosting Machine\n", - "Model Key: best_mgbm\n", - "\n", - "\n", - "Model Summary: " - ] - }, { "data": { "text/html": [ - "
\n", - "\n", - "\n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - "
number_of_treesnumber_of_internal_treesmodel_size_in_bytesmin_depthmax_depthmean_depthmin_leavesmax_leavesmean_leaves
046.046.06939.03.03.03.05.08.07.369565
\n", - "
" - ], - "text/plain": [ - " number_of_trees number_of_internal_trees model_size_in_bytes \\\n", - "0 46.0 46.0 6939.0 \n", - "\n", - " min_depth max_depth mean_depth min_leaves max_leaves mean_leaves \n", - "0 3.0 3.0 3.0 5.0 8.0 7.369565 " + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "
predict p0 p1
10.4315520.568448
10.39618 0.60382
10.4315520.568448
10.4315520.568448
10.3988920.601108
10.4468070.553193
10.39618 0.60382
10.4353710.564629
" ] }, "metadata": {}, "output_type": "display_data" }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "\n", - "\n", - "ModelMetricsBinomial: gbm\n", - "** Reported on train data. **\n", - "\n", - "MSE: 0.13637719864300343\n", - "RMSE: 0.3692928358945018\n", - "LogLoss: 0.4351274080189972\n", - "Mean Per-Class Error: 0.2913939696264273\n", - "AUC: 0.7716491282246187\n", - "pr_auc: 0.5471826859054356\n", - "Gini: 0.5432982564492375\n", - "\n", - "Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.21968260039166268: " - ] - }, { "data": { - "text/html": [ - "
\n", - "\n", - "\n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - "
01ErrorRate
0013482.02814.00.1727(2814.0/16296.0)
111907.02743.00.4101(1907.0/4650.0)
2Total15389.05557.00.2254(4721.0/20946.0)
\n", - "
" - ], - "text/plain": [ - " 0 1 Error Rate\n", - "0 0 13482.0 2814.0 0.1727 (2814.0/16296.0)\n", - "1 1 1907.0 2743.0 0.4101 (1907.0/4650.0)\n", - "2 Total 15389.0 5557.0 0.2254 (4721.0/20946.0)" - ] + "text/plain": [] }, + "execution_count": 10, "metadata": {}, - "output_type": "display_data" - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "\n", - "Maximum Metrics: Maximum metrics at their respective thresholds\n" - ] - }, + "output_type": "execute_result" + } + ], + "source": [ + "best_mgbm.predict(h2o.H2OFrame(poisoned)) # higher scores of model trained on non-poisoned data" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "#### Scores for GBM trained on poisoned data\n", + "When scoring the poisoned data using `poisoned_gbm`, it can be seen that the poisoned data gives surprisingly low probabilities default. If this model is put into production, the adversary can submit similar rows to the poisoned model and expect to receive much lower than normal probabilities of default. These lower probabilities of default could result in the adversary and their associates receiving credit products. It could also result in major financial losses for the credit issuer because formerly high-risk customers could now also receive credit products." + ] + }, + { + "cell_type": "code", + "execution_count": 11, + "metadata": {}, + "outputs": [ { "data": { "text/html": [ - "
\n", - "\n", - "\n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - "
metricthresholdvalueidx
0max f10.2196830.537474248.0
1max f20.1278590.630227329.0
2max f0point50.4466990.583033147.0
3max accuracy0.4466990.821493147.0
4max precision0.9502471.0000000.0
5max recall0.0506091.000000395.0
6max specificity0.9502471.0000000.0
7max absolute_mcc0.3251590.413494194.0
8max min_per_class_accuracy0.1775420.698495281.0
9max mean_per_class_accuracy0.2196830.708606248.0
\n", - "
" - ], - "text/plain": [ - " metric threshold value idx\n", - "0 max f1 0.219683 0.537474 248.0\n", - "1 max f2 0.127859 0.630227 329.0\n", - "2 max f0point5 0.446699 0.583033 147.0\n", - "3 max accuracy 0.446699 0.821493 147.0\n", - "4 max precision 0.950247 1.000000 0.0\n", - "5 max recall 0.050609 1.000000 395.0\n", - "6 max specificity 0.950247 1.000000 0.0\n", - "7 max absolute_mcc 0.325159 0.413494 194.0\n", - "8 max min_per_class_accuracy 0.177542 0.698495 281.0\n", - "9 max mean_per_class_accuracy 0.219683 0.708606 248.0" + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "
predict p0 p1
00.8485070.151493
00.7493570.250643
00.8042640.195736
00.8550280.144972
00.7751980.224802
00.8411190.158881
00.7551730.244827
00.8695880.130412
" ] }, "metadata": {}, "output_type": "display_data" }, { - "name": "stdout", - "output_type": "stream", - "text": [ - "\n", - "Gains/Lift Table: Avg response rate: 22.20 %, avg score: 22.00 %\n" - ] - }, + "data": { + "text/plain": [] + }, + "execution_count": 11, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "poisoned_gbm.predict(h2o.H2OFrame(poisoned)) # lower scores of model trained on poisoned data" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "# Part 2 - Adversarial Examples (Example Exploratory Attack)\n", + "Unlike a data poisoning attack, an adversarial example attack is conducted treating the model as a black box, and only interacting with the predictions of the black box model. In an adversarial example attack, the adversary attempts to learn rows of data that can cause the model to generate the prediction the adversary desires." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "#### Generate random data to score with black box MGBM\n", + "The adversary may have some access to information about the training data such as public documentation or domain knowledge of the features used in the model. Below the adversary uses such knowledge to construct a best guess for what model training data might look like. " + ] + }, + { + "cell_type": "code", + "execution_count": 12, + "metadata": {}, + "outputs": [], + "source": [ + "# best guess at feature distributions\n", + "schema_dict = {'PAY_0': {'mean': 0, 'scale': 1, 'dist': 'normal'},\n", + " 'PAY_2': {'mean': 0, 'scale': 1, 'dist': 'normal'},\n", + " 'PAY_3': {'mean': 0, 'scale': 1, 'dist': 'normal'},\n", + " 'PAY_4': {'mean': 0, 'scale': 1, 'dist': 'normal'},\n", + " 'PAY_5': {'mean': 0, 'scale': 1, 'dist': 'normal'},\n", + " 'PAY_6': {'mean': 0, 'scale': 1, 'dist': 'normal'},\n", + " 'LIMIT_BAL': {'min': 500, 'scale': 1000000, 'dist': 'exponential'},\n", + " 'PAY_AMT1': {'min': 0, 'scale': 80000, 'dist': 'exponential'},\n", + " 'PAY_AMT2': {'min': 0, 'scale': 80000, 'dist': 'exponential'},\n", + " 'PAY_AMT4': {'min': 0, 'scale': 80000, 'dist': 'exponential'}}\n", + "\n", + "N = 10000 # rows of simulated data\n", + "\n", + "random_frame = pd.DataFrame(columns=list(schema_dict.keys())) # init empty frame\n", + " \n", + "for j in list(schema_dict.keys()): # loop through features\n", + " \n", + " np.random.seed(SEED) # same results each time cell is run\n", + " \n", + " # simulate PAY_* features\n", + " if schema_dict[j]['dist'] == 'normal':\n", + " random_frame[j] = np.random.normal(loc=schema_dict[j]['mean'],\n", + " scale=schema_dict[j]['scale'], \n", + " size=N)\n", + " \n", + " # simulate LIMIT_BAL, PAY_AMT* features\n", + " if schema_dict[j]['dist'] == 'exponential':\n", + " random_frame[j] = schema_dict[j]['min'] + np.random.exponential(scale=schema_dict[j]['scale'], \n", + " size=N)\n", + " " + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "#### Calculate partial dependence for each feature in black box MGBM\n", + "Partial dependence can be calculated with **only** model predictions. The adversary will begin the adversarial example attack by calculating partial dependence based on their simulated training data. The knowledge supplied by partial dependence will help narrow the search for adversarial examples." + ] + }, + { + "cell_type": "code", + "execution_count": 13, + "metadata": {}, + "outputs": [], + "source": [ + "# init dict to hold partial dependence and ICE values\n", + "# for each feature\n", + "# for mgbm\n", + "random_pd_ice_dict = {}\n", + "\n", + "# calculate partial dependence for each selected feature\n", + "for xs in list(schema_dict.keys()): \n", + " random_pd_ice_dict[xs] = explain.pd_ice(xs, random_frame, best_mgbm)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "#### Find some percentiles of yhat in the validation data\n", + "ICE will show even more fine-grained details to help select adversarial examples. ICE can be plotted for just one or many individuals. Since no particular individual is known to the adversary, random rows at the deciles of `p_DEFAULT_NEXT_MONTH` are selected for ICE calculations." + ] + }, + { + "cell_type": "code", + "execution_count": 14, + "metadata": {}, + "outputs": [ { "data": { - "text/html": [ - "
\n", - "\n", - "\n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - "
groupcumulative_data_fractionlower_thresholdliftcumulative_liftresponse_ratescorecumulative_response_ratecumulative_scorecapture_ratecumulative_capture_rategaincumulative_gain
010.0100740.8139273.6078833.6078830.8009480.8434460.8009480.8434460.0363440.036344260.788259260.788259
120.0203380.7955753.5198083.5634320.7813950.8051530.7910800.8241190.0361290.072473251.980795256.343177
230.0303160.7636793.4053283.5113940.7559810.7839700.7795280.8109050.0339780.106452240.532798251.139446
340.0400080.7151383.2618913.4509540.7241380.7398150.7661100.7936840.0316130.138065226.189099245.095388
450.0500810.6644163.1168693.3837550.6919430.6866950.7511920.7721640.0313980.169462211.686898238.375473
560.1000190.5433842.8594633.1219840.6347990.6017940.6930790.6871010.1427960.312258185.946339212.198445
670.1500050.3662372.2242932.8228490.4937920.4469510.6266710.6070760.1111830.423441122.429306182.284922
780.2056720.2927651.5955102.4906590.3542020.3127770.5529250.5274220.0888170.51225859.551043149.065864
890.3012510.1966481.1745042.0730770.2607390.2344990.4602220.4344850.1122580.62451617.450421107.307684
9100.4000290.1738170.8643271.7746040.1918800.1848440.3939610.3728420.0853760.709892-13.56728477.460410
10110.5002860.1514310.7014181.5595370.1557140.1613350.3462160.3304550.0703230.780215-29.85824955.953665
11120.6003060.1312140.6192371.4028700.1374700.1407090.3114360.2988410.0619350.842151-38.07634240.286982
12130.7006590.1147940.5593141.2820500.1241670.1228170.2846140.2736300.0561290.898280-44.06856828.204987
13140.8008210.1022260.3692931.1678870.0819830.1080620.2592700.2529210.0369890.935269-63.07069716.788724
14150.9045640.0918610.4021521.0800660.0892770.0975240.2397740.2350990.0417200.976989-59.7848088.006633
15161.0000000.0348100.2411121.0000000.0535270.0769890.2219990.2200100.0230111.000000-75.8887830.000000
\n", - "
" - ], "text/plain": [ - " group cumulative_data_fraction lower_threshold lift \\\n", - "0 1 0.010074 0.813927 3.607883 \n", - "1 2 0.020338 0.795575 3.519808 \n", - "2 3 0.030316 0.763679 3.405328 \n", - "3 4 0.040008 0.715138 3.261891 \n", - "4 5 0.050081 0.664416 3.116869 \n", - "5 6 0.100019 0.543384 2.859463 \n", - "6 7 0.150005 0.366237 2.224293 \n", - "7 8 0.205672 0.292765 1.595510 \n", - "8 9 0.301251 0.196648 1.174504 \n", - "9 10 0.400029 0.173817 0.864327 \n", - "10 11 0.500286 0.151431 0.701418 \n", - "11 12 0.600306 0.131214 0.619237 \n", - "12 13 0.700659 0.114794 0.559314 \n", - "13 14 0.800821 0.102226 0.369293 \n", - "14 15 0.904564 0.091861 0.402152 \n", - "15 16 1.000000 0.034810 0.241112 \n", - "\n", - " cumulative_lift response_rate score cumulative_response_rate \\\n", - "0 3.607883 0.800948 0.843446 0.800948 \n", - "1 3.563432 0.781395 0.805153 0.791080 \n", - "2 3.511394 0.755981 0.783970 0.779528 \n", - "3 3.450954 0.724138 0.739815 0.766110 \n", - "4 3.383755 0.691943 0.686695 0.751192 \n", - "5 3.121984 0.634799 0.601794 0.693079 \n", - "6 2.822849 0.493792 0.446951 0.626671 \n", - "7 2.490659 0.354202 0.312777 0.552925 \n", - "8 2.073077 0.260739 0.234499 0.460222 \n", - "9 1.774604 0.191880 0.184844 0.393961 \n", - "10 1.559537 0.155714 0.161335 0.346216 \n", - "11 1.402870 0.137470 0.140709 0.311436 \n", - "12 1.282050 0.124167 0.122817 0.284614 \n", - "13 1.167887 0.081983 0.108062 0.259270 \n", - "14 1.080066 0.089277 0.097524 0.239774 \n", - "15 1.000000 0.053527 0.076989 0.221999 \n", - "\n", - " cumulative_score capture_rate cumulative_capture_rate gain \\\n", - "0 0.843446 0.036344 0.036344 260.788259 \n", - "1 0.824119 0.036129 0.072473 251.980795 \n", - "2 0.810905 0.033978 0.106452 240.532798 \n", - "3 0.793684 0.031613 0.138065 226.189099 \n", - "4 0.772164 0.031398 0.169462 211.686898 \n", - "5 0.687101 0.142796 0.312258 185.946339 \n", - "6 0.607076 0.111183 0.423441 122.429306 \n", - "7 0.527422 0.088817 0.512258 59.551043 \n", - "8 0.434485 0.112258 0.624516 17.450421 \n", - "9 0.372842 0.085376 0.709892 -13.567284 \n", - "10 0.330455 0.070323 0.780215 -29.858249 \n", - "11 0.298841 0.061935 0.842151 -38.076342 \n", - "12 0.273630 0.056129 0.898280 -44.068568 \n", - "13 0.252921 0.036989 0.935269 -63.070697 \n", - "14 0.235099 0.041720 0.976989 -59.784808 \n", - "15 0.220010 0.023011 1.000000 -75.888783 \n", - "\n", - " cumulative_gain \n", - "0 260.788259 \n", - "1 256.343177 \n", - "2 251.139446 \n", - "3 245.095388 \n", - "4 238.375473 \n", - "5 212.198445 \n", - "6 182.284922 \n", - "7 149.065864 \n", - "8 107.307684 \n", - "9 77.460410 \n", - "10 55.953665 \n", - "11 40.286982 \n", - "12 28.204987 \n", - "13 16.788724 \n", - "14 8.006633 \n", - "15 0.000000 " + "{0: 4999,\n", + " 99: 7436,\n", + " 10: 4026,\n", + " 20: 3419,\n", + " 30: 5370,\n", + " 40: 196,\n", + " 50: 4196,\n", + " 60: 5437,\n", + " 70: 8073,\n", + " 80: 5477,\n", + " 90: 7366}" ] }, + "execution_count": 14, "metadata": {}, - "output_type": "display_data" - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "\n", - "\n", - "ModelMetricsBinomial: gbm\n", - "** Reported on validation data. **\n", - "\n", - "MSE: 0.13326994104124376\n", - "RMSE: 0.3650615578792757\n", - "LogLoss: 0.4278285715046422\n", - "Mean Per-Class Error: 0.2856607030196092\n", - "AUC: 0.7776380047998697\n", - "pr_auc: 0.5486322626112021\n", - "Gini: 0.5552760095997393\n", - "\n", - "Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.27397344199105433: " - ] - }, + "output_type": "execute_result" + } + ], + "source": [ + "# merge MGBM predictions onto random data\n", + "mgbm_yhat_random = pd.concat([random_frame.reset_index(drop=True),\n", + " best_mgbm.predict(h2o.H2OFrame(random_frame))['p1'].as_data_frame()],\n", + " axis=1)\n", + "\n", + "# rename yhat column\n", + "mgbm_yhat_random = mgbm_yhat_random.rename(columns={'p1':'p_DEFAULT_NEXT_MONTH'})\n", + "\n", + "# find percentiles of predictions\n", + "mgbm_percentile_dict = explain.get_percentile_dict('p_DEFAULT_NEXT_MONTH', mgbm_yhat_random, 'index')\n", + "\n", + "# display percentiles dictionary\n", + "# key=percentile, val=Pandas index\n", + "mgbm_percentile_dict" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "#### Calculate ICE curve values" + ] + }, + { + "cell_type": "code", + "execution_count": 15, + "metadata": {}, + "outputs": [], + "source": [ + "# loop through selected features\n", + "for xs in list(schema_dict.keys()): \n", + "\n", + " # collect bins used in partial dependence\n", + " bins = list(random_pd_ice_dict[xs][xs])\n", + " \n", + " # calculate ICE at percentiles \n", + " # using partial dependence bins\n", + " # for each selected feature\n", + " for i in sorted(mgbm_percentile_dict.keys()):\n", + " col_name = 'Percentile_' + str(i)\n", + " random_pd_ice_dict[xs][col_name] = explain.pd_ice(xs, \n", + " pd.DataFrame(random_frame.loc[int(mgbm_percentile_dict[i]), :]).T, \n", + " best_mgbm, \n", + " bins=bins)['partial_dependence']" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "#### View partial dependence and ICE for generated random data and black box MGBM\n", + "Just like a data scientist might use partial dependence and ICE to understand more about a model, an adversary can do the same thing, but use the gained knowledge for destructive purposes." + ] + }, + { + "cell_type": "code", + "execution_count": 16, + "metadata": {}, + "outputs": [ { "data": { - "text/html": [ - "
\n", - "\n", - "\n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - "
01ErrorRate
006093.0975.00.1379(975.0/7068.0)
11863.01123.00.4345(863.0/1986.0)
2Total6956.02098.00.203(1838.0/9054.0)
\n", - "
" - ], + "image/png": "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\n", "text/plain": [ - " 0 1 Error Rate\n", - "0 0 6093.0 975.0 0.1379 (975.0/7068.0)\n", - "1 1 863.0 1123.0 0.4345 (863.0/1986.0)\n", - "2 Total 6956.0 2098.0 0.203 (1838.0/9054.0)" + "" ] }, - "metadata": {}, + "metadata": { + "needs_background": "light" + }, "output_type": "display_data" }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "\n", - "Maximum Metrics: Maximum metrics at their respective thresholds\n" - ] - }, { "data": { - "text/html": [ - "
\n", - "\n", - "\n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - "
metricthresholdvalueidx
0max f10.2739730.549951217.0
1max f20.1478350.634488307.0
2max f0point50.4366200.590736153.0
3max accuracy0.4569630.825271147.0
4max precision0.9470691.0000000.0
5max recall0.0451061.000000397.0
6max specificity0.9470691.0000000.0
7max absolute_mcc0.3472460.429999184.0
8max min_per_class_accuracy0.1815850.709970275.0
9max mean_per_class_accuracy0.2305180.714339240.0
\n", - "
" - ], + "image/png": "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\n", + "text/plain": [ + "" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + }, + { + "data": { + "image/png": "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\n", "text/plain": [ - " metric threshold value idx\n", - "0 max f1 0.273973 0.549951 217.0\n", - "1 max f2 0.147835 0.634488 307.0\n", - "2 max f0point5 0.436620 0.590736 153.0\n", - "3 max accuracy 0.456963 0.825271 147.0\n", - "4 max precision 0.947069 1.000000 0.0\n", - "5 max recall 0.045106 1.000000 397.0\n", - "6 max specificity 0.947069 1.000000 0.0\n", - "7 max absolute_mcc 0.347246 0.429999 184.0\n", - "8 max min_per_class_accuracy 0.181585 0.709970 275.0\n", - "9 max mean_per_class_accuracy 0.230518 0.714339 240.0" + "" ] }, - "metadata": {}, + "metadata": { + "needs_background": "light" + }, "output_type": "display_data" }, { - "name": "stdout", - "output_type": "stream", - "text": [ - "\n", - "Gains/Lift Table: Avg response rate: 21.94 %, avg score: 22.52 %\n" - ] + "data": { + "image/png": "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\n", + "text/plain": [ + "" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" }, { "data": { - "text/html": [ - "
\n", - "\n", - "\n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - "
groupcumulative_data_fractionlower_thresholdliftcumulative_liftresponse_ratescorecumulative_response_ratecumulative_scorecapture_ratecumulative_capture_rategaincumulative_gain
010.0111550.8150103.2950553.2950550.7227720.8398580.7227720.8398580.0367570.036757229.505549229.505549
120.0205430.7955753.7007643.4804600.8117650.8056310.7634410.8242170.0347430.071501270.076417248.045999
230.0300420.7835503.6047213.5197490.7906980.7924410.7720590.8141700.0342400.105740260.472142251.974853
340.0400930.7431923.0058763.3909270.6593410.7613350.7438020.8009250.0302110.135952200.587630239.092657
450.0500330.6977023.4445123.4015730.7555560.7230910.7461370.7854610.0342400.170191244.451158240.157260
560.1012810.5531933.1047773.2513940.6810340.6147360.7131950.6990750.1591140.329305210.477654225.139444
670.1503200.3835642.1870462.9041710.4797300.4660670.6370320.6230610.1072510.436556118.704581190.417123
780.2000220.2969151.5804232.5752440.3466670.3278170.5648810.5496980.0785500.51510658.042296157.524427
890.3013030.2035391.1335142.0906160.2486370.2506480.4585780.4491740.1148040.62990913.351366109.061561
9100.4034680.1769700.9610681.8045950.2108110.1871900.3958390.3828360.0981870.728097-3.89319880.459549
10110.5002210.1520280.6557341.5823820.1438360.1635660.3470960.3404240.0634440.791541-34.42660358.238248
11120.5999560.1330090.5553491.4116510.1218160.1416510.3096470.3073810.0553880.846928-44.46507641.165144
12130.7022310.1150620.4923231.2777570.1079910.1235490.2802770.2806070.0503520.897281-50.76768527.775745
13140.8019660.1023800.3534041.1628020.0775190.1078340.2550610.2591210.0352470.932528-64.65959416.280206
14150.9053460.0918610.3799091.0734050.0833330.0975850.2354520.2406750.0392750.971803-62.0090637.340501
15161.0000000.0348100.2978991.0000000.0653440.0768840.2193510.2251720.0281971.000000-70.2101410.000000
\n", - "
" - ], + "image/png": "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\n", "text/plain": [ - " group cumulative_data_fraction lower_threshold lift \\\n", - "0 1 0.011155 0.815010 3.295055 \n", - "1 2 0.020543 0.795575 3.700764 \n", - "2 3 0.030042 0.783550 3.604721 \n", - "3 4 0.040093 0.743192 3.005876 \n", - "4 5 0.050033 0.697702 3.444512 \n", - "5 6 0.101281 0.553193 3.104777 \n", - "6 7 0.150320 0.383564 2.187046 \n", - "7 8 0.200022 0.296915 1.580423 \n", - "8 9 0.301303 0.203539 1.133514 \n", - "9 10 0.403468 0.176970 0.961068 \n", - "10 11 0.500221 0.152028 0.655734 \n", - "11 12 0.599956 0.133009 0.555349 \n", - "12 13 0.702231 0.115062 0.492323 \n", - "13 14 0.801966 0.102380 0.353404 \n", - "14 15 0.905346 0.091861 0.379909 \n", - "15 16 1.000000 0.034810 0.297899 \n", - "\n", - " cumulative_lift response_rate score cumulative_response_rate \\\n", - "0 3.295055 0.722772 0.839858 0.722772 \n", - "1 3.480460 0.811765 0.805631 0.763441 \n", - "2 3.519749 0.790698 0.792441 0.772059 \n", - "3 3.390927 0.659341 0.761335 0.743802 \n", - "4 3.401573 0.755556 0.723091 0.746137 \n", - "5 3.251394 0.681034 0.614736 0.713195 \n", - "6 2.904171 0.479730 0.466067 0.637032 \n", - "7 2.575244 0.346667 0.327817 0.564881 \n", - "8 2.090616 0.248637 0.250648 0.458578 \n", - "9 1.804595 0.210811 0.187190 0.395839 \n", - "10 1.582382 0.143836 0.163566 0.347096 \n", - "11 1.411651 0.121816 0.141651 0.309647 \n", - "12 1.277757 0.107991 0.123549 0.280277 \n", - "13 1.162802 0.077519 0.107834 0.255061 \n", - "14 1.073405 0.083333 0.097585 0.235452 \n", - "15 1.000000 0.065344 0.076884 0.219351 \n", - "\n", - " cumulative_score capture_rate cumulative_capture_rate gain \\\n", - "0 0.839858 0.036757 0.036757 229.505549 \n", - "1 0.824217 0.034743 0.071501 270.076417 \n", - "2 0.814170 0.034240 0.105740 260.472142 \n", - "3 0.800925 0.030211 0.135952 200.587630 \n", - "4 0.785461 0.034240 0.170191 244.451158 \n", - "5 0.699075 0.159114 0.329305 210.477654 \n", - "6 0.623061 0.107251 0.436556 118.704581 \n", - "7 0.549698 0.078550 0.515106 58.042296 \n", - "8 0.449174 0.114804 0.629909 13.351366 \n", - "9 0.382836 0.098187 0.728097 -3.893198 \n", - "10 0.340424 0.063444 0.791541 -34.426603 \n", - "11 0.307381 0.055388 0.846928 -44.465076 \n", - "12 0.280607 0.050352 0.897281 -50.767685 \n", - "13 0.259121 0.035247 0.932528 -64.659594 \n", - "14 0.240675 0.039275 0.971803 -62.009063 \n", - "15 0.225172 0.028197 1.000000 -70.210141 \n", - "\n", - " cumulative_gain \n", - "0 229.505549 \n", - "1 248.045999 \n", - "2 251.974853 \n", - "3 239.092657 \n", - "4 240.157260 \n", - "5 225.139444 \n", - "6 190.417123 \n", - "7 157.524427 \n", - "8 109.061561 \n", - "9 80.459549 \n", - "10 58.238248 \n", - "11 41.165144 \n", - "12 27.775745 \n", - "13 16.280206 \n", - "14 7.340501 \n", - "15 0.000000 " + "" ] }, - "metadata": {}, + "metadata": { + "needs_background": "light" + }, "output_type": "display_data" }, { - "name": "stdout", - "output_type": "stream", - "text": [ - "\n", - "\n", - "Scoring History: " - ] + "data": { + "image/png": "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\n", + "text/plain": [ + "" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" }, + { + "data": { + "image/png": "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\n", + "text/plain": [ + "" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + }, + { + "data": { + "image/png": "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\n", + "text/plain": [ + "" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + }, + { + "data": { + "image/png": "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\n", + "text/plain": [ + "" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + }, + { + "data": { + "image/png": "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\n", + "text/plain": [ + "" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + } + ], + "source": [ + "for xs in list(schema_dict.keys()): \n", + " explain.plot_pd_ice(xs, random_pd_ice_dict[xs])" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "#### Generate potential adversarial examples\n", + "In the partial dependence and ICE results, it appears that the row at the 90th percentile of `p_DEFAULT_NEXT_MONTH ` has the most natural variance under the model. The adversary will base their search for adversarial examples off this row. The adversary will perturb this row of data thousands of time and submit the perturbed rows to the model to determine their affect on model predictions." + ] + }, + { + "cell_type": "code", + "execution_count": 17, + "metadata": {}, + "outputs": [], + "source": [ + "adversary_frame = pd.DataFrame(columns=list(schema_dict.keys()))\n", + "row = random_frame.iloc[7366, :] # row selected from ICE plots\n", + "\n", + "# search for adversarial examples across four features\n", + "for a in list(random_pd_ice_dict['PAY_0']['PAY_0']): \n", + " for b in list(random_pd_ice_dict['PAY_2']['PAY_2']):\n", + " for c in list(random_pd_ice_dict['LIMIT_BAL']['LIMIT_BAL']):\n", + " for d in list(random_pd_ice_dict['PAY_AMT1']['PAY_AMT1']):\n", + " row['PAY_0'] = a\n", + " row['PAY_2'] = b\n", + " row['LIMIT_BAL'] = c\n", + " row['PAY_AMT1'] = d\n", + " adversary_frame = adversary_frame.append(row, ignore_index=True, sort=False)\n", + "\n", + "# get best_mgbm predictions on adversary_frame\n", + "adversary_frame['p_DEFAULT_NEXT_MONTH'] = best_mgbm.predict(h2o.H2OFrame(adversary_frame)).as_data_frame()[\"p1\"]" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "#### View low scoring adversarial examples \n", + "The adversary now possesses rows of data that can generate almost any desired score from the black box model. Below are rows the adversary could use to generate low probabilities of default to potentially receive a credit product." + ] + }, + { + "cell_type": "code", + "execution_count": 18, + "metadata": {}, + "outputs": [ { "data": { "text/html": [ @@ -1668,534 +1162,100 @@ " \n", " \n", " \n", - " \n", - " timestamp\n", - " duration\n", - " number_of_trees\n", - " training_rmse\n", - " training_logloss\n", - " training_auc\n", - " training_pr_auc\n", - " training_lift\n", - " training_classification_error\n", - " validation_rmse\n", - " validation_logloss\n", - " validation_auc\n", - " validation_pr_auc\n", - " validation_lift\n", - " validation_classification_error\n", + " PAY_0\n", + " PAY_2\n", + " PAY_3\n", + " PAY_4\n", + " PAY_5\n", + " PAY_6\n", + " LIMIT_BAL\n", + " PAY_AMT1\n", + " PAY_AMT2\n", + " PAY_AMT4\n", + " p_DEFAULT_NEXT_MONTH\n", " \n", " \n", " \n", " \n", - " 0\n", - " \n", - " 2020-05-28 14:33:23\n", - " 43.415 sec\n", - " 0.0\n", - " 0.415591\n", - " 0.529427\n", - " 0.500000\n", - " 0.000000\n", - " 1.000000\n", - " 0.778001\n", - " 0.413815\n", - " 0.526105\n", - " 0.500000\n", - " 0.000000\n", - " 1.000000\n", - " 0.780649\n", - " \n", - " \n", - " 1\n", - " \n", - " 2020-05-28 14:33:23\n", - " 43.443 sec\n", - " 1.0\n", - " 0.407822\n", - " 0.511864\n", - " 0.716131\n", - " 0.534717\n", - " 3.474912\n", - " 0.236370\n", - " 0.405538\n", - " 0.507496\n", - " 0.726731\n", - " 0.537125\n", - " 3.444264\n", - " 0.187652\n", - " \n", - " \n", - " 2\n", - " \n", - " 2020-05-28 14:33:23\n", - " 43.467 sec\n", - " 2.0\n", - " 0.401483\n", - " 0.498746\n", - " 0.744646\n", - " 0.532172\n", - " 3.529706\n", - " 0.228731\n", - " 0.398808\n", - " 0.493698\n", - " 0.752909\n", - " 0.534588\n", - " 3.422307\n", - " 0.232825\n", - " \n", - " \n", - " 3\n", - " \n", - " 2020-05-28 14:33:23\n", - " 43.489 sec\n", - " 3.0\n", - " 0.396471\n", - " 0.489013\n", - " 0.748189\n", - " 0.535621\n", - " 3.529706\n", - " 0.228636\n", - " 0.393394\n", - " 0.483273\n", - " 0.756448\n", - " 0.535692\n", - " 3.422307\n", - " 0.214491\n", - " \n", - " \n", - " 4\n", - " \n", - " 2020-05-28 14:33:23\n", - " 43.515 sec\n", - " 4.0\n", - " 0.392442\n", - " 0.481430\n", - " 0.750121\n", - " 0.535358\n", - " 3.529706\n", - " 0.210780\n", - " 0.389030\n", - " 0.475135\n", - " 0.758511\n", - " 0.536095\n", - " 3.422307\n", - " 0.217915\n", - " \n", - " \n", - " 5\n", - " \n", - " 2020-05-28 14:33:23\n", - " 43.535 sec\n", - " 5.0\n", - " 0.389141\n", - " 0.475375\n", - " 0.750058\n", - " 0.535198\n", - " 3.529706\n", - " 0.245059\n", - " 0.385453\n", - " 0.468630\n", - " 0.758505\n", - " 0.535659\n", - " 3.422307\n", - " 0.214270\n", - " \n", - " \n", - " 6\n", - " \n", - " 2020-05-28 14:33:23\n", - " 43.570 sec\n", - " 6.0\n", - " 0.386399\n", - " 0.470332\n", - " 0.756986\n", - " 0.535024\n", - " 3.529706\n", - " 0.243961\n", - " 0.382447\n", - " 0.463157\n", - " 0.764722\n", - " 0.536039\n", - " 3.422307\n", - " 0.229843\n", - " \n", - " \n", - " 7\n", - " \n", - " 2020-05-28 14:33:23\n", - " 43.592 sec\n", - " 7.0\n", - " 0.384191\n", - " 0.466316\n", - " 0.757005\n", - " 0.535418\n", - " 3.529706\n", - " 0.243961\n", - " 0.380045\n", - " 0.458834\n", - " 0.764634\n", - " 0.536411\n", - " 3.422307\n", - " 0.220013\n", - " \n", - " \n", - " 8\n", - " \n", - " 2020-05-28 14:33:23\n", - " 43.614 sec\n", - " 8.0\n", - " 0.382341\n", - " 0.462760\n", - " 0.761106\n", - " 0.540176\n", - " 3.514359\n", - " 0.247446\n", - " 0.378063\n", - " 0.455049\n", - " 0.770340\n", - " 0.542043\n", - " 3.457524\n", - " 0.204330\n", - " \n", - " \n", - " 9\n", - " \n", - " 2020-05-28 14:33:23\n", - " 43.639 sec\n", - " 9.0\n", - " 0.380701\n", - " 0.459589\n", - " 0.762515\n", - " 0.540880\n", - " 3.518279\n", - " 0.235654\n", - " 0.376184\n", - " 0.451464\n", - " 0.772358\n", - " 0.543522\n", - " 3.457524\n", - " 0.223548\n", - " \n", - " \n", - " 10\n", - " \n", - " 2020-05-28 14:33:23\n", - " 43.668 sec\n", - " 10.0\n", - " 0.379202\n", - " 0.456705\n", - " 0.762522\n", - " 0.541424\n", - " 3.518279\n", - " 0.235606\n", - " 0.374583\n", - " 0.448380\n", - " 0.772982\n", - " 0.543893\n", - " 3.457524\n", - " 0.226309\n", - " \n", - " \n", - " 11\n", - " \n", - " 2020-05-28 14:33:23\n", - " 43.697 sec\n", - " 11.0\n", - " 0.378052\n", - " 0.454467\n", - " 0.761648\n", - " 0.541505\n", - " 3.521332\n", - " 0.231023\n", - " 0.373354\n", - " 0.445973\n", - " 0.772925\n", - " 0.544553\n", - " 3.460882\n", - " 0.228960\n", - " \n", - " \n", - " 12\n", - " \n", - " 2020-05-28 14:33:23\n", - " 43.729 sec\n", - " 12.0\n", - " 0.377043\n", - " 0.452420\n", - " 0.762767\n", - " 0.541658\n", - " 3.521332\n", - " 0.229972\n", - " 0.372199\n", - " 0.443670\n", - " 0.773412\n", - " 0.543195\n", - " 3.460882\n", - " 0.224542\n", - " \n", - " \n", - " 13\n", - " \n", - " 2020-05-28 14:33:23\n", - " 43.762 sec\n", - " 13.0\n", - " 0.376137\n", - " 0.450517\n", - " 0.764795\n", - " 0.543264\n", - " 3.525899\n", - " 0.234317\n", - " 0.371369\n", - " 0.441932\n", - " 0.774161\n", - " 0.543632\n", - " 3.448038\n", - " 0.227413\n", - " \n", - " \n", - " 14\n", - " \n", - " 2020-05-28 14:33:23\n", - " 43.796 sec\n", - " 14.0\n", - " 0.375357\n", - " 0.448963\n", - " 0.765145\n", - " 0.543113\n", - " 3.525899\n", - " 0.235654\n", - " 0.370549\n", - " 0.440335\n", - " 0.774176\n", - " 0.543202\n", - " 3.448038\n", - " 0.228076\n", - " \n", - " \n", - " 15\n", - " \n", - " 2020-05-28 14:33:23\n", - " 43.848 sec\n", - " 15.0\n", - " 0.374699\n", - " 0.447543\n", - " 0.766118\n", - " 0.544037\n", - " 3.528417\n", - " 0.233219\n", - " 0.369999\n", - " 0.439161\n", - " 0.774592\n", - " 0.543709\n", - " 3.448038\n", - " 0.228297\n", - " \n", - " \n", - " 16\n", - " \n", - " 2020-05-28 14:33:23\n", - " 43.903 sec\n", - " 16.0\n", - " 0.374098\n", - " 0.446341\n", - " 0.766529\n", - " 0.543896\n", - " 3.560713\n", - " 0.229161\n", - " 0.369390\n", - " 0.437926\n", - " 0.775021\n", - " 0.544851\n", - " 3.424855\n", - " 0.226751\n", - " \n", - " \n", - " 17\n", - " \n", - " 2020-05-28 14:33:23\n", - " 43.949 sec\n", - " 17.0\n", - " 0.373534\n", - " 0.445115\n", - " 0.766312\n", - " 0.544208\n", - " 3.568370\n", - " 0.231452\n", - " 0.368810\n", - " 0.436669\n", - " 0.774927\n", - " 0.545957\n", - " 3.442929\n", - " 0.225425\n", - " \n", - " \n", - " 18\n", - " \n", - " 2020-05-28 14:33:23\n", - " 44.004 sec\n", - " 18.0\n", - " 0.373121\n", - " 0.444171\n", - " 0.766785\n", - " 0.544720\n", - " 3.568370\n", - " 0.229352\n", - " 0.368496\n", - " 0.435909\n", - " 0.775256\n", - " 0.545586\n", - " 3.442929\n", - " 0.226530\n", - " \n", - " \n", - " 19\n", - " \n", - " 2020-05-28 14:33:23\n", - " 44.054 sec\n", - " 19.0\n", - " 0.372722\n", - " 0.443360\n", - " 0.767145\n", - " 0.545059\n", - " 3.568370\n", - " 0.226439\n", - " 0.368047\n", - " 0.435006\n", - " 0.775474\n", - " 0.545922\n", - " 3.442929\n", - " 0.224652\n", + " 58099\n", + " -1.329491\n", + " -1.732135\n", + " 1.11824\n", + " 1.11824\n", + " 1.11824\n", + " 1.11824\n", + " 8.282199e+06\n", + " 574198.935115\n", + " 38664.214167\n", + " 38664.214167\n", + " 0.230445\n", + " \n", + " \n", + " 31451\n", + " -2.537423\n", + " -0.524202\n", + " 1.11824\n", + " 1.11824\n", + " 1.11824\n", + " 1.11824\n", + " 3.313244e+06\n", + " 618367.423894\n", + " 38664.214167\n", + " 38664.214167\n", + " 0.230445\n", + " \n", + " \n", + " 31452\n", + " -2.537423\n", + " -0.524202\n", + " 1.11824\n", + " 1.11824\n", + " 1.11824\n", + " 1.11824\n", + " 3.313244e+06\n", + " 662535.912674\n", + " 38664.214167\n", + " 38664.214167\n", + " 0.230445\n", " \n", " \n", "\n", "" ], "text/plain": [ - " timestamp duration number_of_trees training_rmse \\\n", - "0 2020-05-28 14:33:23 43.415 sec 0.0 0.415591 \n", - "1 2020-05-28 14:33:23 43.443 sec 1.0 0.407822 \n", - "2 2020-05-28 14:33:23 43.467 sec 2.0 0.401483 \n", - "3 2020-05-28 14:33:23 43.489 sec 3.0 0.396471 \n", - "4 2020-05-28 14:33:23 43.515 sec 4.0 0.392442 \n", - "5 2020-05-28 14:33:23 43.535 sec 5.0 0.389141 \n", - "6 2020-05-28 14:33:23 43.570 sec 6.0 0.386399 \n", - "7 2020-05-28 14:33:23 43.592 sec 7.0 0.384191 \n", - "8 2020-05-28 14:33:23 43.614 sec 8.0 0.382341 \n", - "9 2020-05-28 14:33:23 43.639 sec 9.0 0.380701 \n", - "10 2020-05-28 14:33:23 43.668 sec 10.0 0.379202 \n", - "11 2020-05-28 14:33:23 43.697 sec 11.0 0.378052 \n", - "12 2020-05-28 14:33:23 43.729 sec 12.0 0.377043 \n", - "13 2020-05-28 14:33:23 43.762 sec 13.0 0.376137 \n", - "14 2020-05-28 14:33:23 43.796 sec 14.0 0.375357 \n", - "15 2020-05-28 14:33:23 43.848 sec 15.0 0.374699 \n", - "16 2020-05-28 14:33:23 43.903 sec 16.0 0.374098 \n", - "17 2020-05-28 14:33:23 43.949 sec 17.0 0.373534 \n", - "18 2020-05-28 14:33:23 44.004 sec 18.0 0.373121 \n", - "19 2020-05-28 14:33:23 44.054 sec 19.0 0.372722 \n", + " PAY_0 PAY_2 PAY_3 PAY_4 PAY_5 PAY_6 LIMIT_BAL \\\n", + "58099 -1.329491 -1.732135 1.11824 1.11824 1.11824 1.11824 8.282199e+06 \n", + "31451 -2.537423 -0.524202 1.11824 1.11824 1.11824 1.11824 3.313244e+06 \n", + "31452 -2.537423 -0.524202 1.11824 1.11824 1.11824 1.11824 3.313244e+06 \n", "\n", - " training_logloss training_auc training_pr_auc training_lift \\\n", - "0 0.529427 0.500000 0.000000 1.000000 \n", - "1 0.511864 0.716131 0.534717 3.474912 \n", - "2 0.498746 0.744646 0.532172 3.529706 \n", - "3 0.489013 0.748189 0.535621 3.529706 \n", - "4 0.481430 0.750121 0.535358 3.529706 \n", - "5 0.475375 0.750058 0.535198 3.529706 \n", - "6 0.470332 0.756986 0.535024 3.529706 \n", - "7 0.466316 0.757005 0.535418 3.529706 \n", - "8 0.462760 0.761106 0.540176 3.514359 \n", - "9 0.459589 0.762515 0.540880 3.518279 \n", - "10 0.456705 0.762522 0.541424 3.518279 \n", - "11 0.454467 0.761648 0.541505 3.521332 \n", - "12 0.452420 0.762767 0.541658 3.521332 \n", - "13 0.450517 0.764795 0.543264 3.525899 \n", - "14 0.448963 0.765145 0.543113 3.525899 \n", - "15 0.447543 0.766118 0.544037 3.528417 \n", - "16 0.446341 0.766529 0.543896 3.560713 \n", - "17 0.445115 0.766312 0.544208 3.568370 \n", - "18 0.444171 0.766785 0.544720 3.568370 \n", - "19 0.443360 0.767145 0.545059 3.568370 \n", - "\n", - " training_classification_error validation_rmse validation_logloss \\\n", - "0 0.778001 0.413815 0.526105 \n", - "1 0.236370 0.405538 0.507496 \n", - "2 0.228731 0.398808 0.493698 \n", - "3 0.228636 0.393394 0.483273 \n", - "4 0.210780 0.389030 0.475135 \n", - "5 0.245059 0.385453 0.468630 \n", - "6 0.243961 0.382447 0.463157 \n", - "7 0.243961 0.380045 0.458834 \n", - "8 0.247446 0.378063 0.455049 \n", - "9 0.235654 0.376184 0.451464 \n", - "10 0.235606 0.374583 0.448380 \n", - "11 0.231023 0.373354 0.445973 \n", - "12 0.229972 0.372199 0.443670 \n", - "13 0.234317 0.371369 0.441932 \n", - "14 0.235654 0.370549 0.440335 \n", - "15 0.233219 0.369999 0.439161 \n", - "16 0.229161 0.369390 0.437926 \n", - "17 0.231452 0.368810 0.436669 \n", - "18 0.229352 0.368496 0.435909 \n", - "19 0.226439 0.368047 0.435006 \n", - "\n", - " validation_auc validation_pr_auc validation_lift \\\n", - "0 0.500000 0.000000 1.000000 \n", - "1 0.726731 0.537125 3.444264 \n", - "2 0.752909 0.534588 3.422307 \n", - "3 0.756448 0.535692 3.422307 \n", - "4 0.758511 0.536095 3.422307 \n", - "5 0.758505 0.535659 3.422307 \n", - "6 0.764722 0.536039 3.422307 \n", - "7 0.764634 0.536411 3.422307 \n", - "8 0.770340 0.542043 3.457524 \n", - "9 0.772358 0.543522 3.457524 \n", - "10 0.772982 0.543893 3.457524 \n", - "11 0.772925 0.544553 3.460882 \n", - "12 0.773412 0.543195 3.460882 \n", - "13 0.774161 0.543632 3.448038 \n", - "14 0.774176 0.543202 3.448038 \n", - "15 0.774592 0.543709 3.448038 \n", - "16 0.775021 0.544851 3.424855 \n", - "17 0.774927 0.545957 3.442929 \n", - "18 0.775256 0.545586 3.442929 \n", - "19 0.775474 0.545922 3.442929 \n", - "\n", - " validation_classification_error \n", - "0 0.780649 \n", - "1 0.187652 \n", - "2 0.232825 \n", - "3 0.214491 \n", - "4 0.217915 \n", - "5 0.214270 \n", - "6 0.229843 \n", - "7 0.220013 \n", - "8 0.204330 \n", - "9 0.223548 \n", - "10 0.226309 \n", - "11 0.228960 \n", - "12 0.224542 \n", - "13 0.227413 \n", - "14 0.228076 \n", - "15 0.228297 \n", - "16 0.226751 \n", - "17 0.225425 \n", - "18 0.226530 \n", - "19 0.224652 " + " PAY_AMT1 PAY_AMT2 PAY_AMT4 p_DEFAULT_NEXT_MONTH \n", + "58099 574198.935115 38664.214167 38664.214167 0.230445 \n", + "31451 618367.423894 38664.214167 38664.214167 0.230445 \n", + "31452 662535.912674 38664.214167 38664.214167 0.230445 " ] }, + "execution_count": 18, "metadata": {}, - "output_type": "display_data" - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "\n", - "See the whole table with table.as_data_frame()\n", - "\n", - "Variable Importances: " - ] - }, + "output_type": "execute_result" + } + ], + "source": [ + "adversary_frame.sort_values(by='p_DEFAULT_NEXT_MONTH').head(n=3) # 3 lowest scores" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "#### View high scoring adversarial examples\n", + "The adversary now possesses rows of data that can generate almost any desired score from the black box model. Below are rows the adversary could use to generate high probabilities of default to potentially deny someone the credit product." + ] + }, + { + "cell_type": "code", + "execution_count": 19, + "metadata": {}, + "outputs": [ { "data": { "text/html": [ @@ -2217,128 +1277,87 @@ " \n", " \n", " \n", - " variable\n", - " relative_importance\n", - " scaled_importance\n", - " percentage\n", + " PAY_0\n", + " PAY_2\n", + " PAY_3\n", + " PAY_4\n", + " PAY_5\n", + " PAY_6\n", + " LIMIT_BAL\n", + " PAY_AMT1\n", + " PAY_AMT2\n", + " PAY_AMT4\n", + " p_DEFAULT_NEXT_MONTH\n", " \n", " \n", " \n", " \n", - " 0\n", - " PAY_0\n", - " 2794.444824\n", - " 1.000000\n", - " 0.693347\n", - " \n", - " \n", - " 1\n", - " PAY_2\n", - " 307.237366\n", - " 0.109946\n", - " 0.076231\n", - " \n", - " \n", - " 2\n", - " PAY_3\n", - " 215.152893\n", - " 0.076993\n", - " 0.053383\n", - " \n", - " \n", - " 3\n", - " PAY_4\n", - " 155.434448\n", - " 0.055623\n", - " 0.038566\n", - " \n", - " \n", - " 4\n", - " PAY_AMT1\n", - " 127.986313\n", - " 0.045800\n", - " 0.031755\n", - " \n", - " \n", - " 5\n", - " PAY_5\n", - " 127.538628\n", - " 0.045640\n", - " 0.031644\n", - " \n", - " \n", - " 6\n", - " PAY_6\n", - " 102.351601\n", - " 0.036627\n", - " 0.025395\n", - " \n", - " \n", - " 7\n", - " LIMIT_BAL\n", - " 82.432350\n", - " 0.029499\n", - " 0.020453\n", - " \n", - " \n", - " 8\n", - " PAY_AMT2\n", - " 58.934135\n", - " 0.021090\n", - " 0.014623\n", - " \n", - " \n", - " 9\n", - " PAY_AMT4\n", - " 58.858047\n", - " 0.021063\n", - " 0.014604\n", + " 165375\n", + " 3.099595\n", + " 3.502240\n", + " 1.11824\n", + " 1.11824\n", + " 1.11824\n", + " 1.11824\n", + " 607.262272\n", + " 8.580982\n", + " 38664.214167\n", + " 38664.214167\n", + " 0.832092\n", + " \n", + " \n", + " 192717\n", + " 4.307528\n", + " 3.099595\n", + " 1.11824\n", + " 1.11824\n", + " 1.11824\n", + " 1.11824\n", + " 607.262272\n", + " 8.580982\n", + " 38664.214167\n", + " 38664.214167\n", + " 0.832092\n", + " \n", + " \n", + " 172872\n", + " 3.502240\n", + " 1.891663\n", + " 1.11824\n", + " 1.11824\n", + " 1.11824\n", + " 1.11824\n", + " 607.262272\n", + " 8.580982\n", + " 38664.214167\n", + " 38664.214167\n", + " 0.832092\n", " \n", " \n", "\n", "" ], "text/plain": [ - " variable relative_importance scaled_importance percentage\n", - "0 PAY_0 2794.444824 1.000000 0.693347\n", - "1 PAY_2 307.237366 0.109946 0.076231\n", - "2 PAY_3 215.152893 0.076993 0.053383\n", - "3 PAY_4 155.434448 0.055623 0.038566\n", - "4 PAY_AMT1 127.986313 0.045800 0.031755\n", - "5 PAY_5 127.538628 0.045640 0.031644\n", - "6 PAY_6 102.351601 0.036627 0.025395\n", - "7 LIMIT_BAL 82.432350 0.029499 0.020453\n", - "8 PAY_AMT2 58.934135 0.021090 0.014623\n", - "9 PAY_AMT4 58.858047 0.021063 0.014604" + " PAY_0 PAY_2 PAY_3 PAY_4 PAY_5 PAY_6 LIMIT_BAL \\\n", + "165375 3.099595 3.502240 1.11824 1.11824 1.11824 1.11824 607.262272 \n", + "192717 4.307528 3.099595 1.11824 1.11824 1.11824 1.11824 607.262272 \n", + "172872 3.502240 1.891663 1.11824 1.11824 1.11824 1.11824 607.262272 \n", + "\n", + " PAY_AMT1 PAY_AMT2 PAY_AMT4 p_DEFAULT_NEXT_MONTH \n", + "165375 8.580982 38664.214167 38664.214167 0.832092 \n", + "192717 8.580982 38664.214167 38664.214167 0.832092 \n", + "172872 8.580982 38664.214167 38664.214167 0.832092 " ] }, - "metadata": {}, - "output_type": "display_data" - }, - { - "data": { - "text/plain": [] - }, - "execution_count": 7, + "execution_count": 19, "metadata": {}, "output_type": "execute_result" } ], "source": [ - "# load saved best model from lecture 1 \n", - "best_mgbm = h2o.load_model('best_mgbm')\n", - "\n", - "# display model details\n", - "best_mgbm" + "adversary_frame.sort_values(by='p_DEFAULT_NEXT_MONTH').tail(n=3) # 3 highest scores" ] }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [] - }, { "cell_type": "markdown", "metadata": {}, @@ -2348,14 +1367,15 @@ }, { "cell_type": "code", - "execution_count": 28, + "execution_count": 20, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ - "Are you sure you want to shutdown the H2O instance running at http://127.0.0.1:54321 (Y/N)? n\n" + "Are you sure you want to shutdown the H2O instance running at http://127.0.0.1:54321 (Y/N)? y\n", + "H2O session _sid_bf69 closed.\n" ] } ], @@ -2382,7 +1402,7 @@ "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", - "version": "3.6.3" + "version": "3.6.9" } }, "nbformat": 4, diff --git a/rmltk/debug.py b/rmltk/debug.py index c1f194e..b78d491 100644 --- a/rmltk/debug.py +++ b/rmltk/debug.py @@ -1,4 +1,6 @@ import pandas as pd +import numpy as np +import string """ @@ -36,34 +38,42 @@ # represent metrics as dictionary for use later METRIC_DICT = { -#### overall performance -'Prevalence': '(tp + fn) / (tp + tn +fp + fn)', # how much default actually happens for this group -'Accuracy': '(tp + tn) / (tp + tn + fp + fn)', # how often the model predicts default and non-default correctly for this group - -#### predicting default will happen -# (correctly) -'True Positive Rate': 'tp / (tp + fn)', # out of the people in the group *that did* default, how many the model predicted *correctly* would default -'Precision': 'tp / (tp + fp)', # out of the people in the group the model *predicted* would default, how many the model predicted *correctly* would default - -#### predicting default won't happen -# (correctly) -'Specificity': 'tn / (tn + fp)', # out of the people in the group *that did not* default, how many the model predicted *correctly* would not default -'Negative Predicted Value': 'tn / (tn + fn)', # out of the people in the group the model *predicted* would not default, how many the model predicted *correctly* would not default - -#### analyzing errors - type I -# false accusations -'False Positive Rate': 'fp / (tn + fp)', # out of the people in the group *that did not* default, how many the model predicted *incorrectly* would default -'False Discovery Rate': 'fp / (tp + fp)', # out of the people in the group the model *predicted* would default, how many the model predicted *incorrectly* would default - -#### analyzing errors - type II -# costly ommisions -'False Negative Rate': 'fn / (tp + fn)', # out of the people in the group *that did* default, how many the model predicted *incorrectly* would not default -'False Omissions Rate':'fn / (tn + fn)' # out of the people in the group the model *predicted* would not default, how many the model predicted *incorrectly* would not default + #### overall performance + 'Prevalence': '(tp + fn) / (tp + tn +fp + fn)', # how much default actually happens for this group + 'Accuracy': '(tp + tn) / (tp + tn + fp + fn)', + # how often the model predicts default and non-default correctly for this group + + #### predicting default will happen + # (correctly) + 'True Positive Rate': 'tp / (tp + fn)', + # out of the people in the group *that did* default, how many the model predicted *correctly* would default + 'Precision': 'tp / (tp + fp)', + # out of the people in the group the model *predicted* would default, how many the model predicted *correctly* would default + + #### predicting default won't happen + # (correctly) + 'Specificity': 'tn / (tn + fp)', + # out of the people in the group *that did not* default, how many the model predicted *correctly* would not default + 'Negative Predicted Value': 'tn / (tn + fn)', + # out of the people in the group the model *predicted* would not default, how many the model predicted *correctly* would not default + + #### analyzing errors - type I + # false accusations + 'False Positive Rate': 'fp / (tn + fp)', + # out of the people in the group *that did not* default, how many the model predicted *incorrectly* would default + 'False Discovery Rate': 'fp / (tp + fp)', + # out of the people in the group the model *predicted* would default, how many the model predicted *incorrectly* would default + + #### analyzing errors - type II + # costly ommisions + 'False Negative Rate': 'fn / (tp + fn)', + # out of the people in the group *that did* default, how many the model predicted *incorrectly* would not default + 'False Omissions Rate': 'fn / (tn + fn)' + # out of the people in the group the model *predicted* would not default, how many the model predicted *incorrectly* would not default } def get_metrics_ratios(cm_dict, _control_level): - """ Calculates confusion matrix metrics in METRIC_DICT for each level of demographic feature. Tightly coupled to cm_dict. @@ -87,12 +97,11 @@ def get_metrics_ratios(cm_dict, _control_level): for level in levels: for metric in METRIC_DICT.keys(): - # parse metric expressions into executable Pandas statements expression = METRIC_DICT[metric].replace('tp', 'cm_dict[level].iat[0, 0]') \ - .replace('fp', 'cm_dict[level].iat[0, 1]') \ - .replace('fn', 'cm_dict[level].iat[1, 0]') \ - .replace('tn', 'cm_dict[level].iat[1, 1]') + .replace('fp', 'cm_dict[level].iat[0, 1]') \ + .replace('fn', 'cm_dict[level].iat[1, 0]') \ + .replace('tn', 'cm_dict[level].iat[1, 1]') # dynamically evaluate metrics to avoid code duplication metrics_frame.loc[level, metric] = eval(expression) @@ -105,7 +114,6 @@ def get_metrics_ratios(cm_dict, _control_level): def air(cm_dict, reference, protected): - """ Calculates the adverse impact ratio as a quotient between protected and reference group acceptance rates: protected_prop/reference_prop. Prints intermediate values. Tightly coupled to cm_dict. @@ -130,11 +138,10 @@ def air(cm_dict, reference, protected): print(protected.title() + ' proportion accepted: %.3f' % protected_prop) # return adverse impact ratio - return protected_prop/reference_prop + return protected_prop / reference_prop def marginal_effect(cm_dict, reference, protected): - """ Calculates the marginal effect as a percentage difference between a reference and a protected group: reference_percent - protected_percent. Prints intermediate values. Tightly coupled to cm_dict. @@ -164,7 +171,6 @@ def marginal_effect(cm_dict, reference, protected): def smd(valid, x_name, yhat_name, reference, protected): - """ Calculates standardized mean difference between a protected and reference group: (mean(yhat | x_j=protected) - mean(yhat | x_j=reference))/sigma(yhat). Prints intermediate values. @@ -192,5 +198,4 @@ def smd(valid, x_name, yhat_name, reference, protected): sigma = valid[yhat_name].std() print(yhat_name.title() + ' std. dev.: %.2f' % sigma) - return (protected_yhat_mean - reference_yhat_mean) / sigma - + return (protected_yhat_mean - reference_yhat_mean) / sigma \ No newline at end of file diff --git a/rmltk/explain.py b/rmltk/explain.py index 975c1a9..7837af6 100644 --- a/rmltk/explain.py +++ b/rmltk/explain.py @@ -309,7 +309,7 @@ def get_png(model_id): _ = subprocess.call(png_args) -def get_cv_dt(x_names, y_names, frame, model_id, seed_, title): +def get_cv_dt(x_names, y_names, train, model_id, seed_, title, valid=None): """ Utility function to train decision trees. @@ -334,7 +334,10 @@ def get_cv_dt(x_names, y_names, frame, model_id, seed_, title): model_id=model_id) # gives MOJO artifact a recognizable name # train single tree model - tree.train(x=x_names, y=y_names, training_frame=h2o.H2OFrame(frame)) + if valid is not None: + tree.train(x=x_names, y=y_names, training_frame=h2o.H2OFrame(train), validation_frame=h2o.H2OFrame(valid)) + else: + tree.train(x=x_names, y=y_names, training_frame=h2o.H2OFrame(train)) # persist MOJO (compiled Java representation of trained model) # from which to generate plot of tree