版权声明：本文为博主原创文章，未经博主允许不得转载。 https://blog.csdn.net/wds2006sdo/article/details/53769805
这几天刚好有环境，打算学习一下深度学习
看了一圈介绍，发现优达学城的深度学习课程作为入门课程还是不错的
今天看了第一章节的视频，顺便做了任务1
任务1难度不大，按照网站上的说明可以完成下载、打包等工作
Problem 6 没有给代码，下面是我的代码，供大家参考
import numpy as np
import pickle
from sklearn.linear_model import LogisticRegression
print 'start load data!'
data = pickle.load(open('notMNIST.pickle','rb'))
train_features = data['train_dataset']
train_labels = data['train_labels']
test_features = data['test_dataset']
test_labels = data['test_labels']
train_len = len(train_features)
train_features = train_features.reshape(train_len,784)
test_len = len(test_features)
test_features = test_features.reshape(test_len,784)
print 'start train!'
classifier = LogisticRegression()
classifier.fit(train_features, train_labels)
print 'start predict!'
score = classifier.score(test_features,test_labels)
print 'The accuray score is %s' % str(score)
训练集大小：30000，测试集大小：10000
实验结果如下，正确率87.94%，正确率还凑活吧，速度挺慢的。
