Inspiration
Inspired by the urgent need to analyze crime data in a meaningful way, our team wanted to transform raw statistics into insights that could actually help improve public safety and policy planning. We saw potential in applying data science to expose crime patterns and empower preventive actions.
What it does
-Perform in-depth Exploratory Data Analysis (EDA) -Visualize crime trends, hotspots, and high-risk regions -Predict future crime rates using regression and ARIMA -Classify districts into high-crime vs low-crime using machine learning -Create a Crime Risk Index to guide resource allocation -Examine links between cities and crime, as well as seasonal variations
How we built it
-Python (Pandas, Matplotlib, Seaborn) for cleaning, EDA, and plotting -Scikit-learn for classification models and clustering -Statsmodels (ARIMA) for time-series forecasting -GeoPandas for mapping crime risk geographically
Challenges we ran into
-Handling inconsistencies and missing values in historical crime data -Mapping district names to geo-boundaries for visualization -Forecasting with limited time granularity (yearly data only) -Balancing different crime types to create a meaningful risk index
Accomplishments that we're proud of
We did it somehow....
What we learned
-How to extract meaningful insights from government datasets -The value of combining statistical thinking with machine learning -How to deal with real-world data challenges like formatting, bias, and granularity -The importance of clear, impactful visual communication
What's next for us:
Next challenge
Log in or sign up for Devpost to join the conversation.