The Neural Network for Accurate Prediction of Stock Price: Systematic literature review

The prediction of stock prices is crucial for investors to explore the future value related to the company they are willing to invest in. Also, to decide whether to progress with this investment or not. It highly depends on the revenue-generating period to overcome their previous losses. This kind of decision requires it to be fast and accurate simultaneously. Current research aims to find studies on the precise prediction of stock prices with the possession of Neural Network NN by determining the suitable type of NN, hyperparameters, and hybrid models. The methodology applied in the research is a systematic literature review (SLR). An overview of thirteen studies has been systematically reviewed. The selection process of these studies depends on the search protocol (inclusion and exclusion criteria). The finding illustrated that ANN, CNN, and LSTMRNN provide better stock price predictions. Besides, the selection of hyperparameters and hybrid models affects the accuracy of the forecast. The novel perspective of the study is to provide a critical synthesis of previous studies on the accurate models and hyperparameters of neural networks, especially for stock price prediction. Therefore, the researcher's concern about stock price prediction will directly follow the result of the current study for better accuracy.