In this blog, I mainly talk about the clustering in the analysis of the data. Clustering is a technique, in the field of machine learning that aims to group data points that share similarities. There exist approaches to clustering each possessing characteristics and applications.

I followed the data to find out the pattern of the police shooting happened in the different states of U.S. Also, I am looking for the data processing method which will help me to find out the missing values or outliers by imputing them with cluster-specific statistics.  While doing the analysis using clustering technique, I mainly uncover underlying structures in data, aids in data reduction, supports segmentation, classification, and anomaly detection, and finds applications in a wide range of domains.

Some of the common types of clustering which I have gone through in brief to use in my analysis are:

K Means Clustering: K Means stands out as a favoured clustering technique. It groups data into K clusters by assessing their similarity. The algorithm repetitively assigns data points to clusters while adjusting the cluster centers until reaching convergence. This approach works well for clusters that’re distinct and spherical, in shape.

DBSCAN (Density-Based Spatial Clustering of Applications with Noise): Data points are grouped using DBSCAN, or density-based spatial clustering of applications with noise. By locating the core, border, and noise spots, it creates clusters. It is resistant to outliers and can locate clusters of any shape.

Mean Shift Clustering: The mode-seeking algorithm Mean Shift locates cluster centres by repeatedly shifting each data point in the direction of the density function mode. It is helpful for locating patterns in the distribution of the data.

Spectral Clustering: Spectral clustering projects data points into a lower-dimensional space where clustering is carried out using the eigenvalues of a similarity matrix. For non-convex clusters, it is helpful.

In conclusion, I will perform one these in my analysis to unveil the information regarding the shooting happened by the police.