Denver Crime: A Case Study

Project Overview

Overview: Denver, the most populous city in Colorado, has grown considerably since the turn of the millennium. This project provides information regarding crime rates, behavior, and other factors associated with crime in a moderately large US city.
Motivation: This analysis attempts to uncover trends and patterns in crime rates and behaviors in the Denver Metro area over the last 5 years.
Agenda: To understand the different factors affecting crime by asking these questions:
- Do crime rates have a seasonality to them?
- Which neighborhoods are the most dangerous?
- What socioeconomic factors are associated with more crime?
- What are the most common types of crime?

Datasets: Denver Crime, Denver Demographics
- Source: Denver Open Data Catalog
Data collection period: January 2018 to September 2023
Tools, skills, and methodologies:
- Python utilized for exploratory data analysis, data cleaning and wrangling, dataset merging, and cluster analysis.
- Libraries used included pandas, NumPy, matplotlib, seaborn, folium, json, and sklearn.
- Tableau storyboard showcased in-depth analyses and findings.
- GitHub repository for more scripts and other analytical details found here.

After the necessary cleaning, wrangling, and merging of the two datasets, I started with a simple EDA to familiarize myself and attempt to uncover any surface-level trends or patterns.
This quickly revealed to me one main takeaway: crime rates have been increasing.
The logical next question is: why are crime rates going up?

After running a linear regression and a cluster analysis attempting to uncover variables most involved with higher crime rates, I attained inconclusive results.
- Refer to the full Tableau storyboard here.
Using a correlation heatmap, I did find some key socioeconomic variables that showed moderately strong correlations to crime rates.
This in conjunction with crime rates by neighborhood allowed me to create a composite “danger score” which ranked all 78 of Denver’s neighborhoods.

While more work can be done to deepen this analysis and gain a higher resolution picture, I did find some important aspects of crime to help guide citizen behavior.
On top of the neighborhood rankings, I also uncovered the most popular day of the week for crime to occur, as well as the months of the year.
- In addition, the most common types of crime were uncovered and ranked.
For the entire story including recommendations, please refer to the full presentation below (and check out my blog post on this project).