Deconstructing Deep Learning
summary
Covers concrete implementations: a CNN built in Keras/TensorFlow runnable in Google Colab, with five hidden layers, 32 convolutional filters, max pooling and 25% dropout, trained in batches of 256 to about 90% accuracy on Fashion‑MNIST; an LSTM with 32‑dim word embeddings, trained three epochs on IMDB reviews, reporting ~87% accuracy. Explains backprop/gradient descent and shows feature maps, misclassified examples, and practical preprocessing steps that changed results.
AI-written summary · about
chapters (35)
- 0:00Start
- 0:26Agenda
- 1:18Deep Learning Biologically inspired machine learning with multi
- 2:46Artificial Neural Network Structur
- 3:51Deep Artificial Neural Network
- 4:10Artificial Neural Network Node
- 5:12Activation Functions
- 7:00Training an Artificial Neural Netwo
- 9:05Updating Weights via Gradient De
- 10:22Deep Learning 101 Summary
- 11:05How computers «see images
- 13:25Another Approach: Convolutional Fi
- 14:18Applying Convolutional Filters
- 15:34Applying ReLU Activation Functio
- 16:02Dimensionality Reduction via Pooli
- 16:58Feature Identification and Dimensionality
- 17:37Use Case: Fashion MNIST
- 18:03CNN Implementation
- 18:44CNN Architecture Overview
- 22:14Training and Testing the CNN
- 22:39CNN Test Results
- 23:09Some examples of incorrectly classified
- 23:33CNN Summary
- 24:46Recurrent Neural Networks to the Re
- 25:15Unrolling a Recurrent Neural Netw
- 25:54Long Short-Term Memory (LSTM) Net
- 26:11A LSTM Node
- 26:48Use Case : IMDB Reviews
- 27:21Example Positive Review
- 27:35Pre-Processing Dataset
- 28:12Word Embedding for NLP
- 29:00LSTM Implementation
- 30:45Training and Testing the LSTM
- 31:22LSTM Test Results
- 32:55Example of an incorrectly classified re
related talks
54mIntroduction to Understanding Deep Learning
01h 39mGoogle Colaboratory and Python: Deep Learning basics for CV tasks
01h 04mWhat you always wanted to know about Deep Learning, but were afraid to ask
27mDemystifying Deep Neural Networks
01h 29mHands-on introduction to Deep Learning with Keras and Tensorflow
44mMachine Learning Exposed
55mA Tour of Deep Learning With C++
01h 01mA practical guide to deep learning
