AI - Mechanisms

Machine Learning

• systems the ability to automatically learn and improve from experience without being explicitly programmed. 

• Machine learning focuses on the development of computer programs that can access data and use it learn for themselves. 

• categories: 

  1. Supervised machine learning algorithms  

  2. Unsupervised machine learning algorithms 

  3. Semi-supervised machine learning algorithms 

  4. Reinforcement machine learning algorithms 

Deep learning

• subset of machine learning (basically machine learning on steroids) 

• made up of layers of increasing complexity. In involves the machine going through massive amounts of data over and over again until it has reached an acceptable level of accuracy. 

For example, it might categorise something as a dog because it has four legs, but this doesn’t differentiate it from a cat, so the next layer of complexity would involve two eyes and ears and a tail, but this still doesn’t differentiate it from a cat. It would keep doing these iterations until it had reached a layer that had something that would categorise an object as definitely a dog, whether it be the paw shape of the jaw shape. 

10 Amazing Examples Of How Deep Learning AI Is Used In Practice

Neural networks (ANN - Artificial Neuron Network)

• Supporting algorithms of deep learning. 

• Modelled on the way that humans learn and involves labelled data --> ANNs are created by programming regular computers to behave as though they are interconnected brain cells. 

• An artificial neural network has anywhere from dozens to millions of artificial neurons—called units—arranged in a series of layers. The input layer receives various forms of information from the outside world. This is the data that the network aims to process or learn about. From the input unit, the data goes through one or more hidden units. The hidden unit’s job is to transform the input into something the output unit can use. 

Google uses a 30-layered neural network to power Google Photos as well as to power its “watch next” recommendations for YouTube videos. Facebook uses artificial neural networks for its DeepFace algorithm, which can recognize specific faces with 97% accuracy. It’s also an ANN that powers Skype’s ability to do translations in real-time. (source)

• It uses trial and error to start with, making educated guess about what the data contains, and then uses the labels to check if it was right. It will continue to do this over and over until It has reached a suitable level of accuracy. 

Flowchart on how you can work out whether something is using AI or not.