Optimal model description of finance and human factor indices

Economists have conducted research on several empirical phenomena regarding the behavior of individual investors, such as how their emotions and opinions influence their decisions. All those emotions and opinions are described by the word Sentiment. In finance, stochastic changes might occur according to investors sentiment levels. In this study, our main goal is to apply several operational research techniques and analyze these techniques’ accurance. Firstly, we represent the mutual effects between some financial process and investors sentiment with multivariate adaptive regression splines (MARS) model. Furthermore, we consider to extend this model by using distinct data mining techniques and compare the gain in accuracy and computational time with its strong alternatives applied in the analyses of the financial data. Hence, the goal of this study is to compare the forecasting performance of sentiment index by using two-stage MARS-NN (neural network), MARS-RF (random forest), RF-MARS, RF-NN, NN-MARS, and NN-RF hybrid models. Furthermore, we aim to classify the peoples’ feelings about economy according to their confidence levels. Moreover, to forecast the underlying state change of the consumer confidence index (CCI) and to observe the relationship with some macroeconomic data (CPI, GDP and currency rate) at a monthly interval, we apply hidden Markov model (HMM). The aim is to detect the switch between these states and to define a path of these states. We also aim to use volatility models for mainly sentiment index, consumer confidence index, and other indices so that we can get better forecasting results from those datasets.