A Comprehensive View for Providing the Decision on Medicare Data

Abstract

Decision making toward Medicare in the running environment has been taken topmost priority. Processing and understanding of Medicare data are a tricky challenge. The recommended system design carries in various stages to work insight to the problem. In this paper, we use machine learning algorithms for accurate decision on Medicare data. Specifically, the paper analyzes heterogeneous-typed data with the use of hierarchical grouping (HG) mechanismin at the preprocessing phase. Application of metrics like multiple aggregate, grouping and re-indexing simplifies the process in the both the phases. While in the development phase, detection outlier with the perspective of claims provided by the provider given at prior (history). Prior cost acts as good indicator for the decision. Use of statistical-based approach the outlier amount is detected. Random forest (RF) algorithm generates RF trees, and they were able to generate accurate results to choose cost of surgery (disease) from the provided data. Our system is useful to evaluate with reasonably low costs and error free, as demonstrated in experimentation on real-world datasets which are publicly available.