UNSUPERVISED LEARNING: Clustering…

Unsupervised learning is a type of machine learning where algorithms analyze and cluster unlabeled datasets to identify hidden patterns, similarities and structures without predefined categories or even human intervention.
You may ask yourself what the difference between supervised and unsupervised learning is.. Well it is the objectives of each. Concrete and ground-truth labels are necessary for supervised learning in order to train models that respond to highly particular queries whereas unsupervised learning aims at analyzing only features, sorting them into groups without specific tasks
How Unsupervised Learning works;
(i)Unlabeled data input- the algorithm is provided with a huge dataset without pre-defined labels
(ii)Pattern discovery- the algorithm processes the raw data provided, identifying relationships, patterns and structures
(iii)Inference and Organization-the algorithm infers rules and organizes the data provided to it by clustering, association rule and dimensionality reduction
(iv) Insight Generation- The result is a model that offers fresh perspectives, such spotting obscure subgroups in the data or identifying linkages might have been missed.
The three main models used in unsupervised learning are; Clustering Models, Dimensional Reduction Models and Association Rule Model.
Lets have a look into the Clustering Model:
Clustering is a technique in unsupervised learning that divides a collection of data points into groups according to how similar they are. It uses unlabeled data to find underlying structures and patterns in the dataset. The techniques of clustering are listed below:
(i)K-Means Clustering which is an unsupervised machine learning technique that is used to partition a number of observations in to clusters where each observation belongs to the cluster with the nearest mean.
(ii)Hierarchial Clustering is an unsupervised machine learning technique in clustering that creates a hierarchy of nested clusters by combining related data points into clusters. Hierarchical clustering creates a hierarchy of nested clusters by combining related data points into clusters. Hierarchical clustering does not require the number of clusters to be predetermined, unlike K-Means clustering. The two types of Hierarchial clustering are: Agglomerative and divisive clustering.
Clustering is an important technique that is practically applicable in various fields. Based on the data characteristics rather than the process itself, it is key to choose the right clustering method to use.