Evaluation Optimal Prediction Performance of MLMs on High-volatile Financial Market Data

Abstract

The present study evaluates the predictionperformance of the multi-machine learning models (MLMs) onhigh-volatile financial markets data sets since 2007 to 2020. Thelinear and nonlinear empirical data sets are comprised on stockprice returns of Karachi stock exchange (KSE) 100-Index ofPakistan and currencies exchange rates of Pakistani Rupees(PKR) against five major currencies (USD, Euro, GBP, CHF &JPY). In the present study, the support vector regression (SVR),random forest (RF), and machine learning-linear regressionmodel (ML-LRM) are under-evaluated for comparativeprediction performance. Moreover, the findings demonstratedthat the SVR comparatively gives optimal predictionperformance on group1. Similarly, the RF relatively gives thebest prediction performance on group2. The findings of studyconcludes that the algorithm of RF is most appropriate fornonlinear approximation/evaluation and the algorithm of SVR ismost useful for high-frequency time-series data estimation. Thepresent study is contributed by exploring comparativeenthusiastic/optimistic machine learning model on multi-naturedata sets. This empirical study would be helpful for finance andmachine-learning pupils, data analysts and researchers,especially for those who are deploying machine-learningapproaches for financial analysis