The DSN Mart Sales Prediction Hackathon is the qualifying competition for the DSN AI Bootcamp. Participants will work with historical retail sales data from DSN Mart to build a machine learning model that predicts the total sales of a product at a specific store.
DSN Mart operates a chain of stores across Nigeria, ranging from small corner shops to flagship hypermarkets in major urban centers, state capitals, and smaller towns. The challenge aims to help the business understand what drives sales at the product level across different store formats and locations to plan stock, pricing, and investment more intelligently.
What the Challenge Involves Participants are expected to: - Explore and understand the provided product-store sales dataset. - Analyse factors influencing product sales across different stores and locations. - Clean and preprocess the data. - Perform exploratory data analysis (EDA) to identify useful patterns. - Engineer relevant features from available product and store information. - Build and evaluate machine learning models. - Generate predictions for the required test data. - Submit predictions to the competition platform and compete on the leaderboard.
Qualification for the DSN AI Bootcamp This is a qualification hackathon where participants' leaderboard performance serves as one of the inputs considered when selecting candidates for the Bootcamp. While participation does not automatically guarantee selection, high performance and rank contribute significantly to the chances of being selected. Success depends on applying sound data science practices to a real-world business problem, demonstrating the ability to Understand, Analyse, Model, Predict, and Communicate.
Evaluation
The task is a regression problem: for each row in test.csv, predict the value of total_sales. Models are evaluated using Root Mean Squared Error (RMSE), where lower scores are better.