Updated: Jan 5
AI or machine learning is about teaching machines how to learn from data. By using machine learning algorithms to learn from data sets, machines can match or surpass human performance in many tasks. There are three learning approaches: supervised, unsupervised, and reinforcement learning as well as the learning tasks associated with each (Oxford AI Programme, 2022).
All these approaches to machine learning
follow the same fundamental workflow, which consists of four main stages:
Manage data: data is collected, prepared, and split for training and testing.
Train model: the task, features, and algorithms are chosen, and the model is trained.
Evaluate model: the trained model is assessed and improved.
Deploy model: the trained model is deployed for prediction on new data, and the model’s performance is monitored and eventually re-retrained.
Supervised learning refers to a situation where a task has an input variable and an output variable, and an algorithm learns to map the input to the output based on examples (Learned-Miller, 2014:2).
Unsupervised learning is when a machine “learns patterns in the input even though no explicit feedback is supplied” (Russell & Norvig, 2016:694). This approach to machine learning is based on input variables only because there are no output variables in training the AI algorithm.
Reinforcement learning is when a machine is not told how to process the data, or which actions to take, but instead learns by examining the outcomes that follow each behaviour (Sutton & Barto, 2018:1). This type of machine learning is based on learning-by-doing.
Each of these approaches has algorithms that are typically associated with it. Data and the learning approach are critical to machine learning.
AI can perform highly complex problem-solving (such as unravelling intricate cancer diagnoses), but it can also suffer major setbacks (such as the potential for racial discrimination) (Oxford AI Programme, 2022). In the coming months #JungianBitsofInformation will be offering a new service to organisations. An Artificial Intelligence service: to identify opportunities for AI in your organisation and guidance on the ethical considerations to address the common pitfalls of AI with a unique perspective from #analyticalpsychology. Register on the site and be the first to hear about the launch of this new service.