Deep Learning with TensorFlow
summary
Hands-on walkthrough builds classifiers on 28x28 digit images using TensorFlow, starting with a linear model that reaches ~82% accuracy, then a multi‑layer perceptron with ReLU and Adam optimizer up to ~95–97%, and finally a convolutional network with 5x5 kernels, pooling and softmax that yields 98% test accuracy. Code shows training loops, minibatches, TensorBoard summaries, saving weights and a frozen protobuf model, and a simple runtime that multiplies inputs through the graph on GPU to produce predicted labels.
AI-written summary · about
chapters (17)
- 0:00Intro
- 1:27The problem with AI
- 3:21Programming vs Machine Learning
- 8:05Artificial Intelligence
- 9:19Hello World
- 9:57Programming
- 11:12Example Problem
- 16:19Logistic Regression
- 17:31The Matrix
- 19:19Max Index
- 23:22Double Use
- 25:37Thought Experiment
- 31:37Gradient Descent
- 33:35TensorFlow
- 36:45Accuracy
- 43:57Output
- 47:43Deep Learning
related talks
01h 07mDeep Learning with TensorFlow
49mDeep Learning With Python & Tensorflow
56mTensorFlow and Deep Learning without a PhD, Part 1
01h 04mWhat you always wanted to know about Deep Learning, but were afraid to ask
44mTensorflow, Tensorflow, What is it that you see?
01h 29mHands-on introduction to Deep Learning with Keras and Tensorflow
55mTensorflow: Learning Machines
47mDeep Learning: a practical introduction from an Android dev perspective