Home

Welcome to the DC-Crime-Prediction wiki! This is a project still on Progress trying to predict the Hot crime spots in Dc based on publicly available. The Approach we are following as a team and the expected deliverables are presented below:

1. First, we Determined which variables provide the highest correlation with crime incidents in the DC metropolitan area
2. we did cool and Amazing Visualization of incidents in a map for different choices of variables and filters(selected features) using different Heatmaps, kernel density estimate plots to correlate variables. The available data was from 2012-2018 and time series analysis on variables and identification of seasonal effects was conducted. Finally, we Determined some anomalies. The Prediction models Regression (to determine the evolution of patterns over time) Clustering (to determine common patterns across census tracts, Wards or Blocks, etc.) still on progress--------------- Classification (to predict if a given crime is likely to happen in a given time and location)

A notebook including Analysis of crime patterns Models for predicting crime will be in the GitHub for DC lab