AN APPLICATION OF MACHINE LEARNING MODELS IN STOCK PRICE TREND PREDICTION

Abstract

Predictive problems using machine learning models are fundamental and widely applied in various fields related to human life, such as weather forecasting, healthcare, and market prices. These problems focus on predicting the outcomes of events or future values based on historical data values through the construction of predictive models. The research content of this paper focuses on building predictive models on time series data from a stock dataset sourced from the VNINDEX exchange. Through analysis techniques, data preprocessing, parameter selection appropriate for each model characteristic, and the construction and training of machine learning models to predict stock price trends, several representative methods are used, including Autoregressive Integrated Moving Average, Vector Autoregression, Holt-Winters, and Facebook Prophet. Experimental results show that the Facebook Prophet method outperforms the other methods in terms of prediction accuracy and performance.