MitchMatt/Accidents-in-kenya

# Accidents in Kenya - Road Accidents Fatality Prediction ### Group Members 1. Mitch Mathiu 2. Deborah Okeyo 3. Michael Omondi 4. Faith Wanjala ### Project Overview This project analyzes road crashes in Kenya from 2012-2023 and builds a machine learning model to predict the likelihood of fatal outcomes in these crashes. By identifying high-risk factors and patterns, this model can support transportation agencies, public safety departments, and urban planners in developing interventions to reduce road fatalities. ### Main Objectives - **Analyze historical crash data** to uncover patterns associated with fatal accidents. - **Train a machine learning model** to predict the likelihood of fatalities in road crashes. - **Deploy a web-based tool** for real-time fatality risk prediction, enabling users to enter crash details and receive a risk assessment. ### Specific Objectives - Identify critical factors that influence fatality risks (e.g., location, time, weather). - Provide insights to guide road safety campaigns, infrastructure improvements, and policy-making. ### Business Understanding Road traffic accidents in Kenya contribute significantly to public health and safety challenges due to high fatality rates. Reducing these fatalities requires identifying factors that heighten death risks in crashes. This project leverages historical crash data to develop a model capable of estimating fatality probabilities, which can inform targeted safety measures. ### Data Understanding The dataset used in this project includes crash data from Kenya (2012-2023) and is sourced from the **World Bank microdata platform**. Key features in the dataset: - **Crash Date and Time**: Includes the time of day and date of each crash. - **Location**: Geographic details like latitude, longitude, and road name. - **Crash Description Keywords**: Descriptors indicating crash details, such as "fatality," "pedestrian," or "motorcycle." - **Weather and Road Conditions**: Environmental con …