Can you predict what happens to a food delivery order?
In this competition, your goal is to build a machine learning model that predicts the final status of food delivery orders using information about the customer, order, restaurant, payment, delivery distance, delivery time, and courier rating.
Predict the fulfillment_state of each order in the test dataset. The possible outcomes are:
DeliveredDelayedCancelledThis is a multi-class classification problem. Your task is to explore the training data, preprocess the features, build a classification model, and generate predictions for the test data.
Food delivery services generate a large amount of information for every order. Understanding this information can help predict whether an order will be successfully delivered, delayed, or cancelled. In this competition, you will work with a small and beginner-friendly dataset containing everyday food delivery information. You will receive:
client_years)cuisine_category)purchase_amount)travel_range_km)arrival_duration)transaction_type)courier_score)purchase_date)Submissions are evaluated using Accuracy. Higher scores are better. For example, if a model correctly predicts 850 out of 1,000 orders: Accuracy = 85%.
The leaderboard will use a portion of the test data for the public score, while the remaining data will be used for the final private ranking.
Participants are encouraged to start with simple machine learning algorithms and gradually improve their models through data preprocessing and feature engineering. Possible approaches include:
You do not need advanced deep learning techniques to participate.