Performance Analysis of Classification Algorithms

Abstract

Classification issues are crucial in machine learning and data mining. Classification difficulties are utilised in medical diagnostics, bank customer estimation, medicinal investigations and emotion analysis. Many categorisation methods have been created with many different parameter inputs. Using hyperparameter optimisation algorithms, this work aims to improve classification success. The ‘heart and iris’ data sets have been classified using K-nearest neighbour, SVM, decision tree and gradient boost techniques. The hyperparameter optimisation algorithms grid search and random search are applied to these selected classification algorithms. Experiments have shown that using hyperparameter optimisation algorithms improves the performance of all classification algorithms. The best parameter values are shown.