AaronOnosala/Kinshasa-Traffic-Speed-Prediction-For-Yango

# Kinshasa-Traffic-Speed-Prediction-For-Yango * Collaborations: Yango, International Tech Company/ DR. Congo, Zindi Africa ## Project Background Yango, an international tech company renowned for transforming global technologies into local services, aimed to enhance its ride-hailing services in Kinshasa by optimizing route planning and ETA predictions. Efficient route optimization is paramount for reducing travel time and improving user experiences. Yango sought to develop a machine learning model to predict average traffic speeds on major roads in Kinshasa at various times of the day, using traffic data collected in September 2023. This project was crucial for providing accurate ETAs, improving service reliability, and ensuring punctuality for passengers. Efficient route optimization is crucial for reducing travel time and enhancing user experiences in ride-hailing services. This capability would enable Yango to optimize routes, improve estimated time of arrival (ETA) predictions, and enhance overall service reliability. ## Problem Statement The primary objective was to build a robust machine learning model capable of predicting average speeds along major roads in Kinshasa every 15 minutes, using traffic data provided by Yango for September 2023. The model needed to consider time-of-day variations and traffic patterns to enable Yango to optimize routes effectively and enhance overall service quality. ## Approach To achieve this, the project was structured into several key phases: ### Data Preprocessing: - __Data Cleaning:__ Addressed missing values, outliers, and inconsistencies in the traffic dataset to ensure high data quality. - __Feature Extraction:__ Generated relevant features, including time of day, day of the week, weather conditions, road types, and historical traffic patterns. ### Exploratory Data Analysis: + __Visualization:__ Conducted comprehensive visual analysis to understand traffic patterns and identify significant trends and correlation …