Skip to content

Used Machine Learning Classification models such as Logistic Regression, KNN, Decision Tree, and Random Forest to predict arrests. Model has 80% accuracy.

Notifications You must be signed in to change notification settings

rspiro9/Terry-Stop-Arrest-Prediction

 
 

Repository files navigation

Module 3 Final Project - Terry Stop Arrest Prediction

Purpose

Utilize database including information on Terry Stops for a classification analysis to predict if an arrest ended up being made as a final resolution to the police terry stop. This research can be used to help officers better prepare for what to expect during a terry stop based on certain variables that are present.

Data Science Process Used

I leveraged the OSEMN (Obtain, Scrub, Explore, Model, Interpret) process for this project. My notebook is organized to follow this process.

The Dataset

Database containing information on every terry stop made by Seattle police officers (i.e. weapon type, if an arrest was made, reported date, officer precinct, etc.). A terry stop is when a police officer stops a person based on 'reasonable suspicion' that the person may have been involved in criminal activity.

Key Questions Investigated:

  1. Predict if an arrest is likely to occur as a result of a terry stop.
  2. What features are most imprtant in predicting an arrest?
  3. How can this research be used to improve police officer prepardness for terry stops?

About

Used Machine Learning Classification models such as Logistic Regression, KNN, Decision Tree, and Random Forest to predict arrests. Model has 80% accuracy.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Jupyter Notebook 100.0%