The most soft and fluffy intro into Machine Learning and Deep Neural Networks
summary
Step‑by‑step code primer that uses TensorFlow.js in the browser to build, compile and train models with tensors (tf.tensor, reshape, data()) and shows sequential and functional APIs with dense layers, softmax output, optimizer + crossentropy loss + accuracy, fit (epochs, batchSize) and predict. Demonstrates dense MNIST baseline reaching about MNIST ~85% dense, CNN ~92% accuracy, applies CNNs to spectrogram voice commands and segmentation tasks, explains overfitting and remedies like dropout and batch normalization.
AI-written summary · about
chapters (17)
- 0:00Introduction
- 2:16What is machine learning
- 5:51Classification task
- 12:32Neural networks blackbox
- 13:14TensorFlow intro
- 17:54Working with tensors
- 21:05Model training
- 27:22Project structure
- 28:04Dense model for classification
- 30:15Convolutional neural networks
- 32:21Neural network applications
- 34:31Image clustering
- 36:20Visual regression testing
- 38:10Text generation with LSTM
- 41:25Training problems and fixes
- 43:59Inspiration and resources
- 44:40TensorFlow community
related talks
21mA Journey into the World of Machine Learning with TensorFlow.js
55mTensorflow: Learning Machines
34mA Powerful Introduction to Machine Learning with Js
01h 04mWhat you always wanted to know about Deep Learning, but were afraid to ask
44mTensorflow, Tensorflow, What is it that you see?
34mML from a software engineer's perspective
46mMachine learning in the browser using TensorFlow.js
31mIntroduction to Machine Learning in Tensorflow.js