Tesla stock prediction with SVM, decision tree and random forest

Abstract

Stock price prediction is a critical task in financial markets, as it helps traders and investors make informed decisions. This paper compares 3 popular machine learning techniques, which are Support Vector Machines (SVM), Decision Trees, and Random Forests, in predicting stock movements. This paper will evaluate and compare the performance of these models for predicting Tesla's stock price movements by using historical stock data. The 20-day Simple Moving Average (SMA20) is used as a key technical indicator which is combined with the basic stock features such as Open, High, Low, and Volume. The results demonstrate that the SVM model achieved the highest accuracy and recall even when only using the SMA20. However, the SVM model provided a better balance between precision (51.5078%) and recall (93.6404%), achieving the highest F1 score which is 66.4591%. The three models have different focuses and advantages, which also lead to the difference between the test values. In this prediction, the SVM model performs the best, compared to the other two models. With the results of the test, the SVM model is perfect for recognizing price increases in a short time with a balance between precision and recall.