Project Title

An Analysis on Police Shootings to Predict Unjustified Shots

Motivation

The death of George Floyd in the hands of the police sparked massive protests across different cities in America and the world. Caught on camera, the incident showed a subdued George Floyd begging to breathe as a police man placed his knee to his neck for several minutes. This caused outrage particularly in the black community where there is a distrust for the police. The Black Lives Matter (BLM) movement was at the forefront of the protests, in recent years, the BLM has gained popularity for their activism against police brutality against Black-Americans.

The protests were mostly peaceful, however, videos were littered all over twitter showing the police firing rubber bullets, stun guns and tear gas at protesters and journalists to intimidate them. This shocked many Americans who were for the first time witnessing their police’s heavy-handedness in dealing with peaceful protesters.

In recent years, there have been numerous cases involving the fatal shooting of Black Americans by the police. In majority of these cases, the police officers involved in the shooting were let off with a slap on the wrist leaving the victims’ families unsatisfied with the ruling.

Based on this, we decided to conduct an analysis of police shooting incidents in America from 2015 to date, to ascertain if the shootings were justified in a bid to help victims and their families seeking justice to get it. To achieve this, we have used Logistic Regression Classifier to create a pipeline for the model which was already available. The model was tweaked a little bit by removing some features from the training dataset to predict the suspects who their shootings were unjust. Hence, this approach could help the victims, their families as well as NGOs, in their lawsuit against accused police officer(s).

The ‘shooting’ dataset contains data about people that have been shot and killed by the police from 2015 to 2020 such as their name, age, gender and race. Additionally, information is provided about their date of shooting and where it happened, their mental state, if they posed any threat, if they had any weapon on them and if the police was wearing a body camera. You can access the dataset through this link.

Contributing

You can contribute to this README file by editing it. Here, you'll find the project files

License

MIT