A Violent Crime Analysis using Fuzzy C-Means Clustering Approach

Clustering Techniques are the most significant method of grouping data points based on certain similarity. There are two ways in clustering techniques, namely hard and soft clustering. Traditional clustering approaches include grouping of each object to only one cluster. However, there are some cases that each object may belong to multiple partitions. Normally, healthcare and educational data encompass multiple clustering. Such multiple partitioning can be accomplished using overlapping clustering and soft or fuzzy clustering approaches. In this work, Fuzzy C-Means clustering model is applied for multiple clustering based on crime rates. The proposed multiple clustering model is evaluated using USArrests dataset and the results are useful to predict the high possibility of crime incidence by visualizing the crime analysis in various states in US.