Senior associate Stephen Ozanne looks at 'Machine Learning' and gives three reasons as to why machine learning is going to radically change the financial services industry.
Since the 17th century mathematicians have been developing formulas in statistics to help make demographic and economic forecasts for policymaking. More recently, with the arrival of digital computers, more powerful algorithms have been developed, which we now use every day for a wide range of tasks, such as weather forecasting and internet searches. Machine learning has evolved from statistics and involves the development of algorithms that can give computers the ability to learn information, without the information being preprogrammed, in order to solve particular problems.
Generally, there are three different forms of machine learning. In 'supervised learning' the computer is given an example question, the data needed to answer it and the correct answer, similar to learning in a classroom. The computer can then learn the process for answering the question in order to answer similar questions on its own. In 'unsupervised learning' the computer is not told what the correct data is that is needed to answer the question, or what the correct answer is, which the computer has to work out for itself. Lastly, 'reinforcement learning' gives the computer a goal, such as winning a game of chess, and it is given feedback each time it plays the game until it wins. In financial services, where most information is now accessible in a digital form, we are starting to see the use of machine learning techniques in automating tasks, decision-making and compliance, which is likely to rapidly grow.
This article first appeared in EnVoyage Magazine.
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.