# Code Marathon ML 2 — Data

来源：https://www.kaggle.com/competitions/code-marathon-ml-2/data

本地 Meta Kaggle 快照；不是本次实时页面。

# Dataset Description

<p>The required train and test data has been provided in a zip file format (data.zip). The training and test data has been supplied in 'train.csv' and 'test.csv' respectively, the descriptions for both of which have been provided below.  </p>
<p>For making a submission, please generate a CSV file in the same format as 'sample_submission.csv'. The 'identifier' field of each prediction must match with the corresponding 'identifier' of the test sample on which the prediction has been made.</p>
The following image contains the specific format that is to be followed while making submissions.

![sample submission format][1]

<h2>File descriptions</h2>
<ul>
<li><strong>train.csv</strong> - the training set, containing 300 training samples</li>
<li><strong>test.csv</strong> - the test set, containing 20000 testing samples, for which the predictions must be made</li>
<li><strong>sample_submission.csv</strong> - a sample submission file in the correct format</li>
</ul>
<h2>Data fields</h2>
<ul>
<li><strong>identifier</strong> - an index unique to each sample, required for generating predictions in the prescribed format</li>
<li><strong>f0, f1, f2, ..., g149 </strong> - the 300 features for each sample</li>
<li><strong>target (only present in 'train.csv')</strong> - the binary target variable (either 0 or 1) that must be predicted</li>
</ul>


  [1]: https://gdurl.com/I5FW