A MACHINE LEARNING APPROACH IN FINANCIAL MARKETS

Abstract

Financial market, a very vital and large sector offinance, includes a large number of investors, buyers and sellers. Financial market trends prediction has been a phenomenon since machine learning was discovered and introduced. But very few approaches became useful for forecasting the financial market as it changes with the passage of time. The objective of this project is to develop a software which can predict and forecast the financial market trends and stocks with the help of machine learning approach. Time Series approach has been used in our project in order to forecast the trend ofthe items. Time series have many models and we have to fit the appropriate model in order to forecast. Then, Seasonal & Trend Decomposition by Loess Forecasting (STLF) is used as it has been widely applied in monetary and financial sectors for its extraordinary and great efficiency for financial market prediction. The STLF Model have two main components which is Trend & Seasonality. Trend means an increase or decrease in the quantity of data & Seasonality means any specific season or time ofthe year. This model works when the combination oftrend and seasonality occurs. We got some accurate results through this model which is very helpful for the investors in order to invest money on those products which are in higher demand.