TensorFlow and Deep Learning without a PhD
summary
Shows how to use TensorFlow to build and train a neural network that recognizes handwritten digits (likely MNIST). Covers practical neural-network design techniques and optimization “tricks of the trade” to raise the model’s recognition accuracy above 99%.
AI-written summary · about
chapters (13)
- 0:00Introduction
- 1:26From array to neuron
- 8:02Training and cross entropy
- 11:22Practical demo
- 14:20TensorFlow program structure
- 22:06Single-layer limits
- 22:50Multi-layer networks and ReLU
- 26:25Learning rate and overfitting
- 28:14Regularization techniques
- 30:06Convolutional networks for images
- 36:17CNN implementation in TensorFlow
- 38:4499% accuracy achieved
- 40:05Resources and conclusion
related talks
56mTensorFlow and Deep Learning without a PhD, Part 1
01h 06mDeep Learning with TensorFlow
01h 07mDeep Learning with TensorFlow
01h 04mWhat you always wanted to know about Deep Learning, but were afraid to ask
49mDeep Learning With Python & Tensorflow
01h 29mHands-on introduction to Deep Learning with Keras and Tensorflow
34mML from a software engineer's perspective
44mTensorflow, Tensorflow, What is it that you see?