Tools, Classification with Neural Nets, PyTorch Implementation

Explore neural network classification and PyTorch implementation in this comprehensive lecture. Learn about tools like Typora, Notion, and Draw.io for visualization. Dive into neural network training, classification techniques, and space-fabric stretching concepts. Understand fully connected layers, inference processes, loss functions, gradient descent, and back-propagation. Follow along with hands-on PyTorch implementations for classification and regression using Jupyter notebooks. Discover the 5-step training process in PyTorch and gain insights into regression uncertainty estimation. Enhance your deep learning skills with practical examples and visual explanations throughout this informative session.

Syllabus

– Welcome!
– Typora
– Notion
– Lecture begins
– Draw.io and inference
– Neural nets training, classification
– Space-fabric stretching animation
– Drawing time! blackboard 2-100-2-5 diagram
– Training data
– Fully connected layer
– Inference
– Training → loss function
– Training → gradient descent & back-propagation
– PyTorch classification implementation with Jupyter notebook
– PyTorch 5-step training
– PyTorch regression implementation with Jupyter notebook
– Regression uncertainty estimation