Beyond Supervision

Abstract

This chapter investigates the domain of unsupervised learning algorithms, delivering a detailed outline of its classifications and essential characteristics. Each algorithm is examined, assessing its appropriateness for different types of data. A systematic assessment is conducted with each algorithm and is checked using datasets that complement its strengths. The evaluation presents insights on how well these algorithms perform in comparison with contextually relevant datasets. It also provides the foundation for a systematic investigation into the details of unsupervised learning by matching each algorithm with an appropriate class of data. The knowledge gap between theory and practice was explored by clustering algorithms with prominent tools like Weka and goes beyond the clustering concept by establishing essential principles for unsupervised learning. Through the explanation of clustering algorithms with real world datasets, practical approaches were provided to employ unsupervised learning in real-world data.