I introduced a new optimized and efficient computational framework in microRNA and their target predictions and easy to use by biological scientists, cost-effective service hosting platform using the application of machine learning. Model training is based on Hierarchical artificial learning, which is also known as Deep learning, although able to provide controlled efficiently fast performance as compared to other published algorithms. The proposed architecture utilizes a combination of statically provisioned dedicated resources and widely available opportunistic public resources to provide quality assured services. My research has led to important discoveries that have advanced in understanding of microRNA and their target as a biomarker in diseases.
I explored the combination in two different platforms and applications. In the first case, the thesis is based on data mining of big data by using different algorithm and literature survey for prediction of accurate interaction and filtering out weed data that leads to lower the accuracy to make it ready to use for Advanced deep learning model. In later case, a combination of dedicated clusters of computers and idle capacities of high-end computers has been exploited to build a classifier framework to serve to compute-intensive applications with response time guarantees.
In conclusion, I have observed that by designing an appropriate framework from different accurate resource algorithms comparing with Benchmark datasets increases the accuracy of interactions, which is state of art technology in personalized medicine.