Written by Machine Learning is a subset of AI that can be considered narrow AI whereby the machine focuses on a specific task.
Essentially machine learning works with data sets to identify patterns and learns from the patterns that it recognises.
Although the pre history of machine learning can be traced all the way back to Thomas Bayes’s work on Bayes Theorem in 1763, modern concept of machine learning can be traced to the 1950s with Turing’s Learning Machine in 1950, the work on Genetic Algorithms, Marvin Minsky and Dean Edmonds Neural machine in 1951 and Arthur Samuel’s chequers game playing machine 1952.
It can be split across:
Supervised
Most real world machine learning utilises supervised learning. The process entails inferring a function from labeled training data.
Semi Supervised
A class of machine learning techniques that utilises a combination of labeled and unlabeled data for training. It is standard practice to use a limited portion of labeled data and a large portion of unlabeled data
Unsupervised
Unsupervised learning is a type of machine learning algorithm used to draw inferences from datasets consisting of input data without labeled responses. The most common unsupervised learning method is cluster analysis.
Reinforcement Learning
An area of machine learning inspired by behaviourist psychology, concerned with how software agents ought to take actions in an environment so as to maximize some notion of cumulative reward.
With machine learning techniques it is important to consider issues such a a clean data set for training and also to check for overfitting. These are topics for a future blog.
A non exhaustive list of use cases for machine learning today include:
- Fintech & banking with risk management, fraud detection (anomaly detection);
- Insurtech with automation of claims using visual images to assess damage and enhancing risk assessment in underwriting in the insurance sector;
- Retail and ecommerce with recommendation algorithms;
- Martech with personalised targeting algorithms for tailored marketing campaigns;
- Health care with diagnostic tools for detection of disease and working with big data sets for readmission prediction helping drive health tech solutions