Comparative study of machine learning algorithms in data classification

Tan, Kai Jun (2025) Comparative study of machine learning algorithms in data classification. Final Year Project, UTAR.

PDF Download (2424Kb)

Abstract

This research conducts a comparative study of various machine learning algorithms for dataset classification to identify the most accurate and reliable classifier. In many different fields, data mining, the process of identifying significant patterns in historical data, is essential to decision-making. One of the medical datasets pertaining to disease diagnosis and prediction, within, diabetes, heart disease, and breast cancer prediction, heart disease are used in this study to assess machine learning classifiers. Logistic regression, decision trees, k-Nearest Neighbors, Naïve Bayes, support vector machines, and random forest approaches are among the classifiers that were examined. The performance of these classifiers will be assessed using key evaluation metrics such as accuracy, precision, recall, f1-score, confusion matrix, AUC-ROC, and precision-recall. The implementatn and comparative analysis will be conducted using Python, and Streamlit. Python scripts can now be transformed into interactive web apps through the integration of Streamlit, enabling dynamic viewing and real-time interaction with classification results. In order to help choose the best algorithm for medical dataset, this study attempts to shed light on the advantages and disadvantages of each classifier. The results will help with real-world data mining applications of machine learning and be a useful guide for further study and practical applications.

Actions (login required)