Improved_RFM_Customer_Segmentation

Improving the RFM model in customer segmentation

Big-Data-Analytics-of-Crimes-in-Uk

The project aims to provide a detailed analysis as well as answer some questions about crimes in the UK. The dataset comprises of approximately 40 million records of crime data from 2010 to 2018. Apache Spark through python (Pyspack) was majorly used.

Heart-Disease-Predictiction

The aim of the project is to hasten heart disease diagnosis by about 30% to save doctors and medical practitioners time. Various classifier models were built with health dataset (about 70,000 records). The models were evaluated with accuracy and Mathews Correlation Coefficient (MCC). Prior to model building, the given dataset was analysed and some interesting insights were uncovered.