Analysis of Machine Learning Models for Road Traffic Accidents
- Led and implemented an end-to-end Machine Learning project pipeline to classify Accident Severity from a highly imbalanced dataset.
- Scrutinized four predictive models - highlighting Random Forest and k-Nearest Neighbors as top performers.
- Boosted the model accuracy up to 84% by incorporating hyper-parameter tuning.
- Evaluated the models using Cross-Validation, and performed Feature Importance to make informed conclusions.
Aesha Shah
ex-Data Engineer Intern @ Amazon | Computer Science
Machine Learning and Data Science enthusiast.