Health economics using feature selection algorithm and regression method for prediction of ending cash
Background: Health economics are amongst academic fields which can aid in ameliorating conditions so as to perform better decisions in regards to the economy such as determining cash prices. The prediction of ending cash is fundamental for internal and external users and can come quite handy in terms of health economics. The most important purpose of financial reporting is the presentation of information to predict ending cash. Ergo, the aim of the research is to predict ending cash value using feature selection and MLR method from 2010-2012. Methods: A feature selection algorithm (Best-First, Greedy-Stepwise and Ranker) was employed in this research to nominate relevant data that affect ending cash. Results: Based on the results of the deployed feature selection method, the following features were indicated as the most relevant in terms of determine ending cash: interest payments for loans, dividends received from short and long term deposits, total net flow of investment activities, net increase (decrease) in cash and beginning cash based on best-first (CFS-Subset-Evaluation) and Greedy-Stepwise (CFS-Subset-Evaluation). Net out flow, dividends, dividends paid, interest payments for loans and dividends received deposits for short and long term were the most important data as indicated by the Ranker (Info-Gain-Attribute-Evaluation, Gain-Ratio-Attribute-Evaluation and Symmetricer-Attribute-Evaluation). According to Ranker (Principal-Components and Relifef-FAttribute-Evaluation) the best data for determining ending cash include beginning cash, interest payments for loans, dividends, net increase (decrease) in cash and dividends received from short and long term deposits. The findings were also indicative of a positive and highly significant correlation between dividends received from short and long term deposits and beginning cash (1.00**), with a significance level of 0.01, whereas the observed correlation between interest payments for loans and ending cash (0.999**), at a significance level of 0.01 was negatively significant. Conclusions: The present research attempted to reduce the volume of data required for predicting end cash by means of employing a feature selection so as to save both precious money and time.