Bear attacks, while rare, have the potential to be fatal if unprepared. Many factors can change the outcome of a bear attack. In this project we hope to find out how fatal a bear attack is based on different factors, such as the species of bear, and specific demographic on the victims. We want to predict what species of bear produced the most fatal attacks.
Our dataset can be found on Kaggle. It is a compilation of data on fatal bear attacks in North America and Canada, from the year 1990 to 2019 . The file itself contains information such as date of attack, location, with specific latitude and longitude of the attack, species of bear, and the victim/victims demographic information, such as sex and age. In addition, our dataset includes a description of such attacks.
Link to dataset: https://www.kaggle.com/datasets/danela/fatal-bear-attacks-north-america
For this project, we’d like to create a logistic regression model, that takes as input variables like a bear species, etc, and outputs a probability of fatality. With this, we’d like to study the coefficients of such model, across multiple bear species to identify the most lethal of them. In addition, we’re interested in studying what are the most lethal states when it comes of bear attacks.
Luis Jimenez, Lucas Rodriguez, Emmanuel Mejia