ISSN:
2324-7657
Clustering crime data helps to identify and understand the underlying crime patterns. The k-means algorithm is very familiar and effectively applied in clustering problems, and grieves from a few drawbacks due to its choice of initial cluster centroids. A hybrid method based on combining the k-means algorithm, improved grasshopper optimization algorithm, and genetic algorithm called KM-IGOA-GA is proposed in this research work. KM-IGOA-GA searches for cluster centers of the dataset as does the k-means algorithm and it can effectively find the global optima. The new KM-IGOA-GA is applied to the crime dataset acquired from NCRB, India and its performance is compared with those of KM-IGOA, KM-GOA, KM-FOA, KM-MFO, and K-means clustering. Results show that KM-IGOA-GA is effective and suitable for data clustering with an accuracy of 89%. This proposed work is implemented using R-Studio.