Forecasting Daily Stock Movement Using a Hybrid Normalization Based Intersection Feature Selection and ANN

• DOI: 10.1016/j.procs.2023.01.121
• Publication Date: 2023-01-31T00:51:26Z
• Authors: Binita Kumari; Tripti Swarnkar
• Keywords: Stock (firearms); Normalization

The potential financial benefits of stock market forecasting have drawn a lot interest. Due to the various interconnected aspects, predicting these markets is difficult endeavour that necessitates thorough as well efficient feature selection procedure discover highest useful aspects. Stock price changes are also influenced by previous trading days' movements, which time series problem. In forecasting, techniques commonly used, although most known systems use single methodology probably can neglect some key notions about regression function at root problem relating variables for input and output. This study employs an artificial neural network (ANN) based generative model forecast pricing in future combining features preferred different picking strategies build ideal optimal group. We begin calculating expanded set 83 technical indicators using day-to-day data six indices, then we normalize them Hybrid-Normalization (HN) technique. important selected types considering common movement prediction. For trend predictions, used variety classifiers such Support Vector Machine, K Nearest Neighbour Artificial Neural Network and. system was given performance review after simulations were done on 6 indices from portions international market. outcomes show joining highlighted got choice calculations taking care into profound beats best class techniques.