Kendall's Coefficient of Concordance Ranking of the Effectiveness of Single Machine Learning Models in Predicting Stock Price Movement
Year: 2020
Type: article
Abstract: The prediction of stock price behavior is extremely important in the world of finance. Generally, stock investment decisions are made by investors by predicting the future movement of prices of stocks. An accurate prediction of the movement of stock price is needed by investors to make decisions regarding buying and selling of stocks, and minimize the risk associated with such investment. In this study, we adopt Kendall's coefficient of concordance technique to rank the effectiveness of five single machine models (which are Decision Tree (DT), Gaussian Naïve Bayes (GNB), Logistic Regression (LR), Support Vector Machine (SVM) and k-Nearest Neighbor (kNN)) in forecasting the direction of movement of stock price. The experiment is conducted utilizing ten different stock data sets randomly collected from three stock markets. Each data set is split into training and test sets. The models are trained on the training data set, and the test data set is used to evaluate the models. The models are evaluated using the six classical evaluation metrics. The experimental results recorded illustrate that LR model had the highest rank when using Kendall's coefficient of concordance (W) with accuracy, precision, f1- score, specificity, and AUC metrics. However, SVM model achieved the best rank when using the Kendall's coefficient of concordance (W) with recall metric.
Authors: Ampomah Ernest Kwame, Qin Zhiguang, Eric Ahudey, Addo Prince Clement
Language: English
Field: Decision Sciences
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W3094166324