ABSTRACT
This chapter examines some key tasks in unsupervised learning such as problem identification, data preprocessing, hyperparameters settings and performance evaluations. It demonstrates unique features of unsupervised machine learning as it is capable of creating a flexible and versatile way to unveil hidden features in data owing to its labeled features independency. The chapter begins by locating all frequently occurring itemsets, and then, by combining them, it develops rules. The algorithms are based on the category of unsupervised learning techniques. Dimensionality reduction is a potent unsupervised learning technique that may be used to streamline complex datasets and enhance the reliability of clustering. It is especially helpful when the data are noisy or have outliers and can't be effectively clustered using more conventional techniques like k-means or hierarchical clustering. Some of the key steps in achieving optimal unsupervised learning processes. External validation approaches are not typically employed on clustering problems because unsupervised learning techniques are typically used when such information is unavailable.
