皮肤疾病诊断 Skin Disease Diagnosis

该项目是人工神经网络课程的课程设计。需要使用Python语言以及pytorch等Python机器学习库，训练一种神经模型来实现对40类皮肤病图片分类任务。

项目难点：

1. 提供的数据集Skin40是一个小样本的数据集，很容易出现过拟合等问题，难以直接使用进行训练；

2. 部分皮肤疾病之间较为相似，存在分类上的困难。

---

This project is the course design of artificial neural network course. It is necessary to use Python language and Python machine learning libraries such as pytorch to train a neural model to realize the task of classifying 40 types of skin disease pictures.

Project difficulty:

1. The data set provided (Skin40) is a small sample data set, which is prone to overfitting and other problems, and is difficult to be directly used for training.

2. Some skin diseases are similar to each other, and there are difficulties in classification.