Traffic accidents severity can be predicted based on previous accidents. From the given example data set of traffic accidents, we can build a model that predict future accidents' severity. This will help drivers make their decision to reduce accidents.
Data source is from Kaggle: https://www.kaggle.com/tbsteal/canadian-car-accidents-19942014?select=drivingLegend.pdf (Open database License).
This dataset describes Canadian Car Accidents 1994-2014 with details. Target information is accidents is either fatal or non-fatal, accompanied by 21 attributes that closely describes the particular cases. Data and meta data are in the repository.
Data studied and ML algorithms deployment can be found in notebook (*.ipynb) or directly at IBM Cloud: https://dataplatform.cloud.ibm.com/analytics/notebooks/v2/e97b21ca-4023-4723-a41a-d8f501919ad9/view?access_token=287e5c54a0e44c12f5af15ea14cf2970b6bf27a5ea0d07b6026c19d7aaf54d83
Data is from Kaggle and project is the Capstone Project of IBM Data Science course, hosted by Coursera. All this repository content is under MIT License.