初学机器学习，第一步是做一个简单的手写数字识别，我选用的是MNIST数据集（用其他数据集也可以，原理都差不多），算法是KNN（下载库直接调用函数，算法的具体实现没有过多关心）。在网上也看到过MNIST数据集的Python代码，但是感觉有些复杂，作为初学者见到那么多代码就头大……这里分享一下我的代码，虽然并不完善，但是可以为其他初学者提供一点简单的思路吧。(Learning machine learning, the first step is to do a simple handwritten numeral recognition, I use the MNIST data set (can also be used, other data sets are similar, the principle of the algorithm is KNN (download) library directly call the function, the algorithm implementation is not too much to care about). The Python code of the MNIST data set has also been seen online, but it feels a bit complicated, and it's too much to see as much code as a beginner...... Here to share my code, although not perfect, but you can provide a simple idea for other beginners.)