Galaxy Morphology Classification Dataset

星系形态分类数据集

1、包含大中小三个数据集，每个数据集都打好了标签，可以直接使用。

2、大中小三个数据集的测试集都是相同的，方便测试训练出的模型。

3、三个数据集的训练集样本数由大到小分别是26153、8295、2868。

4、三个数据集的训练集样本量大约是三倍的关系，便于做同一模型在不同样本数下的表现的对比实验。

5、样本的标签用数字代替，具体标签如下：

00000： 圆形星系 00001： 中间星系 00002： 雪茄状星系 00003： 侧向星系 00004： 漩涡星系

Galaxy morphology classification dataset

1. It contains three data sets, large and small. Each data set is labeled and can be used directly.

2. The test sets of the three data sets are the same, which is convenient for testing the trained model.

3. The sample numbers of the training sets of the three data sets are 26153, 8295 and 2868 respectively from large to small.

4. The sample size of the training set of the three data sets is about three times, which is convenient for the comparison experiment of the performance of the same model under different sample numbers.

5. The labels of the samples are replaced by numbers. The specific labels are as follows:

00000: Round galaxy

00001: Intermediate galaxy

00002: Cigar galaxy

00003: Lateral galaxy

00004: Spiral Galaxy