Modeling Crime Data

By Hannah Jiang


Law enforcement services depend on technology to save lives. Analyzing crime data can reveal patterns among dates and locations, which emergency dispatchers rely on to assign emergency units and minimize response times. Using historical Seattle crime data, this project will use statistical and computational skills to determine when and where crimes occur the most in Seattle, in addition to why this could be happening and how we can predict crime rates in the future. There are three components to this activity, in addition to some of the topics we’ll cover in each:

  • Data Analysis - What types of crime have been most prevalent? Where has the most crime occurred in the last 12 years? How has this changed over time?

  • Building a network - What causes crime? How do these causes relate and how can we visualize them?

  • Machine Learning- How can we predict crimes rates in the future? How could we improve those predictions?

For this activity, the user is required to have Jupyter Notebook installed, since we’ll be using Python and popular packages such as pandas, matplotlib, seaborn, numpy, networkx, and scikit learn.

You'll need to open Github to get started- click the button above and read the "Getting Started" section!

The content of these pages was created by students for students with the help of educators and scientists. The views expressed herein are those of the authors and do not necessarily reflect the views of NSF or ISB.