Predicting stock market using machine learning: best and accurate way to know future stock prices

• Published in: International journal of system assurance engineering and management Vol. 14; no. 1; pp. 1 - 18
• Main Authors: Sheth, Dhruhi, Shah, Manan
• Format: Journal Article
• Language: English
• Published: New Delhi Springer India 01-02-2023; Springer Nature B.V
• Subjects: Artificial intelligence; Artificial neural networks; Engineering; Engineering Economics; Kernel functions; Logistics; Machine learning; Marketing; Neural networks; Organization; Pattern recognition; Pricing; Quality Control; Reliability; Review Paper; Safety and Risk; Securities markets; Support vector machines; Stock market; Machine learning (ML); Prediction
• ISSN: 0975-6809; 0976-4348
• DOI: 10.1007/s13198-022-01811-1

Dissatisfaction is the first step of progress, this statement serves to be the base of using Artifcial Intelligence in predicting stock prices. A great deal of people dreamed of predicting stock prices faultlessly but it remained only as a dream for those visionaries at that time. The legacy of those visionaries led to the discovery of something concrete and made that dream come to reality, and due to this we can use machine learning methods in today’s era for predicting accurate stock prices. These methods have proved to be extremely beneficial and an easy way for common man to earn quick money if done appropriately. These methods still have drawbacks that are being worked upon and it confirmations immense improvement in the future unlike the prior methods of predicting stock market prices like time-series forecasting that didn’t provide results that satisfying the needs of an investor. As a result, to deal with the volatile and dynamic nature of the market, a link between stock market and Artificial Intelligence was founded that brought about wonders. The three methods that were implemented in the prediction process were Artificial Neural Network (ANN), Support Vector Machine (SVM) and Long Short-Term Memory (LSTM). ANN works on neural network, SVM works using Kernel method and LSTM works using Keras LSTM. Various techniques offered by each methodology are carefully analyzed and it was found that ANN based on neural network provides best results because it considers complex, non-linear relationships and recognizes patterns. While SVM is comparatively a new method and capable of providing better results in the future and LSTM gives good results only when large dataset is given which can be considered a drawback.