DATA ANALYTICS - PROJECT

FORECASTING CRIME - BURGLARY USING ARIMA

Contributors:

• Nikhil KR (PES2201800044)
• Roshan Daivajna (PES2201800372)
• Ruchira R Vadiraj (PES2201800602)

Description:

• The link to the dataset can be found here (https://catalog.data.gov/dataset/crimes-2001-to-present-398a4).
• The data is first acquired from the given link and is pre-processed.
• The PartI.py file handles data acquisition, data pre-processing and exploratory data analysis.
• The refined processed data is saved as a CSV file (output.csv - Download locally).
• The second part of the analysis is done in R due to the packages available and ease of use.
• The PartII.R file gets the output.csv file and model selection and forecasting is done.
• The final result obtained is shown as the output, with the error rates.

NOTE :

• Forecasting is done on a particular type of crime, Burglary.
• Data aggregation and further model selection is done in R due to easy data handling provided by R dataframes.