Primer on artificial intelligence
This section of the website is not ready for use
Supervised Learning — A strategy that involves a teacher that is smarter than the network itself. In this situation, the network is trained by providing a set of inputs, and the correct response that should be provided for each input. The network then adjusts it’s prediction matrix based on how correct/incorrect it was.
Unsupervised Learning — Required when there isn’t an example data set with known answers. In this situation the network is attempting to identify patterns on its own through clustering, i.e. dividing a set of elements into groups according to some unknown pattern.
Reinforcement learning - A strategy built on observation. Where supervised learning aims to teach the network through lots of examples, reinforcement learning aims to teach the network through experience. It involves having the network attempt a solution and then obtain feedback (a reward score) that indicates how successful it was.
Types of artificial intelligence
AI engineers love data! We devour it to train our networks. Here are some excellent links to various datasets and pre-trained models.
- Datasets curated by DeepLearning
- Datasets curated by Kaggle
- Datasets curated by Wikipedia
- Discover open source deep learning code and pretrained models @ ModelZoo
- Easy to use and pre-trained machine learning models @ ModelDepot
- A topic-centric list of HQ open datasets in public domains @ awesomedata on github
Valuable resources I highly recommend: