AI-900: AI Fundamentals
Types of Learning
Supervised Learning
- Labelded data is used for training
- Task-driven - make a prediction
- eg. classification, regresssion
Unsupervised Learning
- Unlabelled data is used for training
- Data-driven - recognize a pattern
- e.g. clustering, association
Reinforcement Learning
- No data; an environment and an ML model that generates data to reach a goal
- Decision-driven
- e.g. Game AI, Robot Navigation
Neural Network
- Algorithms trying to mimick the brain
- Data is input into a neuron - which represents an algorithm and the output is passed to one of many other connected neurons
- The connection between neurons is weighted and the network is organized in layers - 1 input layer, 1 to many hidden layers, 1 output layer
- A neural network that has 3 or more hidden layers is considered deep learning
GPU vs. CPU
General Processing Unit
- Designed to quickly render high-resolution images and videos concurrently
- Have thousands of processor cores
- Optimized for parallel operations on multiple sets of data
Central Processing Unit
- Designed to handle a wide range of tasks quickly
- Have on average 4-16 processor cores
- Optimized for serial operations
Computer Vision
- Use neural networks to gain high-level understanding of digital images or videos
- Convolutional neutral network (CNN)
- Recurrent neural network (RNN)
____
2 November 2022
Share this post: