Full-Batch Dataset

该数据集用于训练神经网络模型，其中包含了经过多种优化算法处理的全批量数据，旨在达到最高的训练准确率。此外，实验采用了具有三层结构、每层含有100个隐藏单元和ReLU激活函数的全连接神经网络。该数据集还考虑了不同的随机种子（从0到9）进行实验，以扩大神经网络训练任务的规模和多样性。

This dataset is intended for training neural network models, and it encompasses full-batch data processed via multiple optimization algorithms, with the goal of achieving the highest possible training accuracy. Furthermore, the experiment adopts a fully connected neural network with a three-layer architecture, where each layer contains 100 hidden units and ReLU activation functions. Additionally, this dataset uses various random seeds ranging from 0 to 9 in experiments to expand the scale and diversity of neural network training tasks.