Ground-truth dataset for XAI

该数据集是一份精心设计的合成数据集，其中包含了明确界定的线性统计依赖关系。它旨在评估不同的解释性人工智能（XAI）方法在区分重要特征与非重要特征方面的能力。数据集包含了基于不同信噪比的变化，并且被划分为训练集和验证集。规模上，该数据集包含100个子集，每个子集有1000个样本。任务是对数据集中的统计依赖关系进行二分类，以区分两个不同的类别。

This is a meticulously designed synthetic dataset containing well-defined linear statistical dependencies. It is developed to assess the performance of various eXplainable Artificial Intelligence (XAI) methods in discriminating critical features from non-critical ones. The dataset incorporates variations across different signal-to-noise ratio (SNR) levels and is split into training and validation subsets. In terms of scale, the dataset consists of 100 subsets, with each subset containing 1000 samples. The task is a binary classification problem targeting the statistical dependencies within the dataset, aiming to distinguish between two distinct classes.