# Connected Workshop: Surrogate Model Challenge — Data

来源：https://www.kaggle.com/competitions/atelier-connecte-surrogate-model-challenge/data

本地 Meta Kaggle 快照；不是本次实时页面。

# Dataset Description

📂 Files

train.csv – training set. It contains input variables, customer demands and workshop parameters, and target KPIs (WIP, investment, satisfaction).

test.csv – test set. It contains only input variables, customer demands and parameters. It does not include KPIs.

sample_submission.csv – example submission file. It shows the expected format for predictions.

solution.csv (hidden) – used only by Kaggle to evaluate predictions. It contains the true KPI values.

📑 Main columns

id – unique identifier for a scenario.

feature_1 … feature_N – variables that describe customer demands, for example helicopter type, number of units, and technical characteristics.

param_1 … param_M – variables that describe the workshop configuration, for example number of operators, production lines, and available resources.

wip – target KPI. Work In Progress.

investment – target KPI. Cost or invested resources.

satisfaction – target KPI. Customer satisfaction rate.

⚠️ Reminder

Train your models on train.csv.

Generate predictions on test.csv.

Submit results using the exact format of sample_submission.csv.