https://drive.google.com/file/d/1dk3QxkZklfpMX4hnV-OU52mTFVwc56Vt/view?usp=share_link
Inspiration
We were inspired by the small number of variables in the dataset which necessitated a create approach.
What it does
Our analyses describe the trends by time and location of every crime type in Montreal from 2015-2023 and give statistically significant, quantified odds of danger depending on time of day, year, and location for different types of crime.
How I built it
EDA was performed using Python and statistical analyses were done in R.
Challenges I ran into
The dataset was not conducive to logistic regression and a massive overhaul was needed to fit this model. Also, feature engineering was particularly difficult due to all the different datatypes in the set.
Accomplishments that I'm proud of
W performed every type of analysis we wanted to and didn't have to compromise on quality.
What I learned
Daytime is the most dangerous time in the city overall, but nighttime in the summer is particularly bad.
What's next for Montreal Crime Analysis
Use our models to predict future crimes.
Log in or sign up for Devpost to join the conversation.