Machine Intelligence - Lecture 13 (Convolutional Neural Networks, CNNs)

Показать описание

SYDE 522 – Machine Intelligence (Winter 2019, University of Waterloo)

Target Audience: Senior Undergraduate Engineering Students

Course Outline - The objective of this course is to introduce the students to the main concepts of machine intelligence as parts of a broader framework of “artificial intelligence”. An overview of different learning, inference and optimization schemes will be provided, including Principal Component Analysis, Support Vector Machines, Self-Organizing Maps, Decision Trees, Backpropagation Networks, Autoencoders, Convolutional Networks, Fuzzy Inferencing, Bayesian Inferencing, Evolutionary algorithms, and Ant Colonies.

Lecture 13 - Convolutional Neural Networks, CNNs

Рекомендации по теме

Комментарии

Discrete convolutions, weighted sums and many fast transforms are dot products. Max pooling is switching. ReLU is a switch 🤔 f(x)=x connect, f(x)=0 disconnect. A light switch in your house is binary on off yet connects and disconnects a continuously variable AC voltage signal. A ReLU net then is a switched composition of dot products. The dot product of a number of dot products is still a dot product. Then you can see that when all the switch states in a ReLU net become known it collapses to a simple matrix, with various things you can look at.
You can swap what is adjusted in a neural network. You can have fixed dot products (enacted with fast transforms) and adjustable (parametric) activation functions like fi(x)=ai.x x<0, fi(x)=bi.x x>=0, i=0 to m. The fast Walsh Hadamard transform can be viewed as a system of dot products where the cost per dot product is a very low log2(n) add subtracts. Such a Fast Transform fixed filter bank neural network then is: transform, functions, transform, To stop the first transform from taking the spectrum of the input data (which is a bit unhelpful) you can apply a fixed randomly chosen (or sub-random) pattern of sign flips to the input to the net. Such a net likewise collapses to a simple matrix for a particular input. There is a theme here😀

hoaxuan

0:00 - 0:04 who's that cute boy on the bottom left 😍😍

silverrapth

Machine Intelligence - Lecture 13 (Convolutional Neural Networks, CNNs)

Machine Intelligence - Lecture 13 (Convolutional Neural Networks, CNNs)

Machine Intelligence - Lecture 18 (Evolutionary Algorithms)

AI & ML in Finance - Lecture 13 - Elastic Net & Cross-Validation

MIT: Machine Learning 6.036, Lecture 13: Clustering (Fall 2020)

Machine Intelligence - Lecture 3 (PCA, AI and Data)

Lecture 13: Introduction to Convolutional Neural Networks (CNN) – Machine Learning for Engineers

Ethics of Artificial Intelligence - Part 1 :: Machine Intelligence Course, Lecture 23

Lecture 13 | AI Free Basic Course

Machine Intelligence - Lecture 19 (Opposition-Based Learning, GAs, DE)

Machine Intelligence - Lecture 14 (Overfitting in Deep Learning, Reinforcement Learning)

Machine Intelligence - Lecture 5 (Computer Vision, Features, Fisher Vector, VLAD)

EfficientML.ai Lecture 13 - Transformer and LLM (Part II) (MIT 6.5940, Fall 2023)

Lecture 13: Convolutional Neural Networks

Lecture 13 - How to be a Great Founder (Reid Hoffman)

CS480/680 Lecture 13: Support vector machines

Machine Intelligence - Lecture 2 (Turing Test, Chinese Room, Generalization, PCA)

Machine Learning - Lecture 13

Machine Intelligence - Lecture 15 (Reinforcement Learning, Q-Learning)

Lecture 13: ML Teams (Full Stack Deep Learning - Spring 2021)

Lecture 13 | Generative Models

PIM Course: Lecture 13: Bit-Serial SIMD Processing using DRAM - Fall 2022

Lecture 13: Bayes Nets

13. Learning: Genetic Algorithms

Machine Intelligence - Lecture 17 (Fuzzy Logic, Fuzzy Inference)