TensorFlow

Use TensorFlow to create neural network

use jupyter to open ipynb

non-line_regression:

#梯度下降

train_step=tf.train.GradientDescentOptimizer(0.1).minimize(loss)

CNN-MNIST.ipynb:

用了CNN后，第一轮train的数字识别准确率就达到95%。

Final Result of Handwritten digit recognition:

用图形化显示 “手写数字”分类的结果

Name		Name	Last commit message	Last commit date
Latest commit History 29 Commits
.ipynb_checkpoints		.ipynb_checkpoints
MNIST_data		MNIST_data
logs		logs
projector		projector
screenshoot		screenshoot
CNN-MNIST.ipynb		CNN-MNIST.ipynb
Kmeans.ipynb		Kmeans.ipynb
MNISTSimpleVersion.ipynb		MNISTSimpleVersion.ipynb
README.md		README.md
SVM.ipynb		SVM.ipynb
non-line_regression.ipynb		non-line_regression.ipynb
tensorboard.ipynb		tensorboard.ipynb
tensorshow.py		tensorshow.py
tf-deep(MNIST_improvement).ipynb		tf-deep(MNIST_improvement).ipynb
总结.txt		总结.txt

Provide feedback