Neural nets training
This lecture covers the concept of neural nets training (tools, classification with neural nets, and PyTorch implementation) and is a part of the Deep Learning Course at CDS, a course that covered the latest techniques in deep learning and representation learning, focusing on supervised and unsupervised deep learning, embedding methods, metric learning, convolutional and recurrent nets, with applications to computer vision, natural language understanding, and speech recognition. Prerequisites for this course include: Introduction to Data Science or a Graduate Level Machine Learning.
Chapters:
00:00 – Welcome!
00:45 – Typora
01:27 – Notion
03:12 – Lecture begins
04:48 – Draw.io and inference
12:01 – Neural nets training, classification
18:00 – Space-fabric stretching (animation)
20:26 – Drawing time! (blackboard 2-100-2-5 diagram)
26:06 – Training data
32:07 – Fully connected layer
39:06 – Inference
42:11 – Training → loss function
47:14 – Training → gradient descent & back-propagation
50:00 – PyTorch classification implementation with Jupyter notebook
53:03 – PyTorch 5-step training
1:01:29 – PyTorch regression implementation with Jupyter notebook
1:03:40 – Regression uncertainty estimation