yickysan/omdena-kenya-task-5-web-development

# Predicting Road Accident Severity This is a project which I built while collaborating with Omdena Kenya local chapter. My task was to build a machine learning web app and deploy a machine learning model for predicting road accidents. ## Project Goals * This project aims to leverage machine learning (ML) techniques to analyze and predict road accidents in Kenya. * Analyze and understand the patterns and contributing factors of accidents on Kenyan roads using * historical accident data. * Identify accident-prone areas (hotspots) by analyzing accidents' spatial and temporal patterns. * Develop a machine learning model to predict the severity of accidents based on various factors such as road conditions, weather, time of day, and vehicle types. ## Requirements * pandas * numpy * plotly * matplotlib * seaborn * scikit-learn * notebook * flask ## Run To run the project you need to install the dependencies using `pip install -r requirements.txt` This is going to build the OmdenaKenyaRoadAccidents package in your local machine. There are 5 major components: * DataIngestion class * DataTransformation class * ModelTrainer class * train_pipeline function * PredictPipeline class These components are used to download the data from Google drive, apply preprocessing, add a machine learning model and then make predictions from the form input coming from the web app. The DataIngestion class `initiate_data_ingestion` method is used to download the data from Google drive. This method takes one argument `id` which is the file id of the data to be downloaded. The downloaded file is the split into a train and test data set, with the test data being 20% of the original dataset. The datasets are stored in the artifacts directory. #### data ingestion example ```python from OmdenaKenyaRoadAccidents.components.data_ingestion import DataIngestion id = "sYHT1jdjdieiW?ejieX3" data_ingestion = DataIngestion() train_path, test_path = data_ingestion.initiate_data_ingestion(id=id) ``` The …