This paper discusses the long-standing problem of household segregation within the black area and its correlation with crime rates. Despite the progress in the area of socioeconomic improvement, domestic isolation continues to be one of the major obstacles faced by African Americans. The study aims to predict crime rates accurately, and it investigates the use of data techniques, particularly those developed from the UCI repository1 . African Americans experience a higher degree of racial profiling within the penal system than whites. Given the increasing crime magnitude all over the globe, analyzing crime data is an urgent necessity to reduce crime and speed up how cases are solved. The paper uses a recursive feature elimination technique for feature selection through a dataset containing around 2000 records and 128 fields. The Multiple Linear Regression(MLR) method is used for feature selection, followed by the analysis using the Logistic Regression Algorithm (LRA). The proposed model is applied to the "Communities and Crime Data Set" from UCI Machine Learning Datasets to achieve a remarkable accuracy of 99.92%. This implies the success of the approach towards forecasting or prediction of crime patterns and emphasizes the significance of data-driven models in dealing with critical issues within communities, like crime. 1https://archive.ics.uci.edu/dataset/183/communities+and+crime