# HackOWASP ML Contest - Discontinued — Data

来源：https://www.kaggle.com/competitions/hack-owasp-ml-contest/data

采集时间：2026-10-19T21:36:47Z


## Dataset

The dataset consists of two main files:

* **train.csv:** This file contains the training data, including a unique `id`, 100 features (`feature_1` to `feature_100`), and the target variable (`Target`).
* **test.csv:** This file contains the test data, including a unique `id` and the same 100 features (`feature_1` to `feature_100`). Participants will make predictions for the `Target` variable for each `id` in this file.

The format of the data is comma-separated values (CSV).

**Column Descriptions:**

* **id:** A unique identifier for each data point.
* **feature\_1** to **feature\_100:** One hundred numerical features used for prediction. The specific nature and scale of these features will be evident in the data.
* **Target:** The continuous target variable that needs to be predicted. This column is only present in the `train.csv` file.

## Evaluation

The evaluation metric for this competition is **R-squared** (Coefficient of Determination). Higher R-squared values indicate a better fit of the model to the data.

$$R^2 = 1 - \frac{SS_{res}}{SS_{tot}}$$

where $SS_{res}$ is the sum of squared residuals and $SS_{tot}$ is the total sum of squares.

## Submission Format

The submission file should be a CSV file named `submission.csv` containing two columns: `id` and `target`. The `id` column should contain the same unique identifiers from the `test.csv` file, and the `target` column should contain your model's predictions for the corresponding `id`. The file must include a header row.

```csv
id,target
1,0.32
2,10.33
3,5.6
```