🎯 Model Accuracy Hard

Campus Cafeteria Traffic Prediction

Problem Description

You are given 30 days of hourly cafeteria foot traffic data (7:00–21:00, 15 hours/day), along with features:

  • Date, day of week
  • Whether it's a holiday
  • Weather condition (sunny/cloudy/rainy)
  • Temperature (°C)
  • Actual visitor count

Your task: Predict hourly traffic for days 31–37 (next week).

Submit a CSV with 105 rows (7 days × 15 hours) containing your predictions.

Training Data Overview

Last 7 days of training data shown. Clear daily patterns with lunch peaks.

Scoring Rubric

Your Submission

Upload your predictions CSV

105 rows: day 31–37, hours 7–21

Quantitative Metrics

RMSE

lower is better

R² Score

higher is better

Peak Hour MAE

lunch 11-13h

Prediction vs Reference

Your prediction Reference Error region

Error by Day × Hour

Hour
Absolute error: < 20 20–50 > 50

Overall Evaluation

60% quantitative + 40% qualitative

/100

Overall Assessment

💡 Improvement Suggestions