Business Problem: The Financial Officers of the Healthcare providers have insufficient information to understand the reasons for unpredictable long revenue cycle.
Our Solution: The work done was to predict the cash flow from claim reimbursements for healthcare providers using historical data. The predictions made were
- How much money can be expected?
- When is the money expected?
- Likelihood of claim payment
A ground-up design, development and implementation of a Predictive model was provided. A proprietary hybrid model was developed using a home grown scoring model and a Support Vector Machine (SVM) model. The model involved an active classification mechanism for the claims and an auto-model training mechanism. The outcome was presented in an interactive dashboard with highly intuitive graphics.
Technology : R 2.14.2 with RServe, RODBC, RPART e1071, SQL Scripts, Oracle 11g, Dashboard built using LDF with custom Flex components