Data Mining and Visualisation for Business Intelligence Project Report

This data mining project utilizes the Weka workbench to analyze and evaluate the performance of various classification algorithms across three datasets: diabetes.arff, iris.arff, and breast cancer.arff. The project begins with data preprocessing, followed by the implementation and evaluation of five classification algorithms: MultilayerPerception, Naive Bayes, J48, RandomForest, and REPTree. The evaluation includes detailed accuracy metrics such as correctly classified instances, Kappa statistic, mean absolute error, and confusion matrices for each algorithm and dataset. The results are presented in a structured format, providing insights into the strengths and weaknesses of each algorithm. The project demonstrates the application of data mining techniques for classification and performance evaluation, which are essential in business intelligence. The project also includes detailed results from each dataset, demonstrating the practical application of the algorithms and their comparative performance, enabling the student to understand the data mining process comprehensively.