Artificial Intelligence (Al) is far from a new concept. Cars, cell phones, fraud detection and prevention technology, ad recommendations that pop on the side of your Facebook page, etc. are all forms of technology that utilize artificial intelligence in one form or another. Within the larger field of AI, there are various subsets such as robotics, computer vision, and perhaps most famously machine learning (ML). Machine learning is a sophisticated extension of Al that is rapidly advancing the efficiency of the modern world, and in fact, you probably use ML-powered technology on a daily basis.
Coined by A.L. Samuel in 1959, machine learning is a process whereby a computer system learns and adapts to data it is fed, essentially training itself. Instead of humans programming the technology to do or know something, the technology is programmed to teach itself based on its experiences
With supervised learning, the model is provided with a known, desired outcome. The model then trains itself to best achieve that outcome. The challenge, of course, is then feeding new data to the existing model. Often, the model is “overfit” on the data that it was trained on and is useless in making decisions with new data.
In unsupervised learning, the model is not provided with any target outcome. Often, unsupervised learning is used to identify similar clusters, or groups, of observations. Recommendation systems are a famous use case of unsupervised learning.
Finally, reinforcement learning is quickly gaining momentum. In this approach, the model learns from feedback it is given throughout the learning process. For example, when training a dog to sit, the dog is rewarded with a treat once it performs the task it is commanded to do. After the dog is praised and rewarded for its accomplishment, the dog will know to do the same thing when asked to execute the task again. Reinforcement learning uses the same idea, but without the dog treat, of course.
You may be wondering what an every-day example of Machine Learning is. The truth is that ML is embedded in multiple forms of technology that you more than likely use regularly. For example, the spam filter in your email is one common application of ML. Based off what you have opened and responded to in the past, ML will sort what it thinks you will find important from unimportant. It is important to consider that the more information it is fed, the more accurate it will be.
The views expressed represent the opinion of Passage Global Capital Management, LLC. The views are subject to change and are not intended as a forecast or guarantee of future results. This material is for informational purposes only. It does not constitute as investment advice and is not intended as an endorsement of any specific investment. Stated information is derived from proprietary and nonproprietary sources that have not been independently verified for accuracy or completeness. While Passage Global Capital Management, LLC believes the information to be accurate and reliable, we do not claim or have responsibility for its completeness, accuracy, or reliability. Statements of future expectations, estimates, projections, and other forward-looking statements are based on available information and Passage Global Capital Management, LLC’s views as of the time of these statements. Accordingly, such statements are inherently speculative as they are based on assumption that may involve known and unknown risks and uncertainties. Actual results, performance or events may differ materially from those expressed or implied in such statements.