Understanding Supervised vs Unsupervised Machine Learning Models

Machine learning is a hot topic in almost all fields of study and application - and crypto is no exception! Have you ever read the terms "Supervised machine learning" or "unsupervised machine learning" and thought: "Does this mean that a data scientist is supervising the computer crunch numbers?!" How exactly is a machine learning algorithm "supervised" or not? Does the unsupervised model mess up more than the supervised model? When would we want to use one type of model vs another? These are the sort of questions that we will tackle in this article.

Supervised vs Unsupervised Models
Let's get right down to it. The most basic way to understand the difference between these two broad categories of machine learning hinges on: whether or not the correct result is "known" beforehand. For example, If we have a bunch of images that are cats, and these images constitute our "training set" of data for our model, then this would qualify as a supervised learning model as the "correct" answers are known before we train the model. (We know which images are "Cats" and which are not beforehand) Unsupervised models group categorical data by similarities and feature dimensions, where the "correct" grouping is not known before the algorithm is run.

Does an unsupervised model make more mistakes than a supervised model?
Short answer: no. Well, not necessarily. Both models rely on the quality of the feature set, or the categorical (or numerical) data to make predictions. And the features of your model that you use may not have any sort of predictive power, rendering both models useless. This follows the old "garbage in, garbage out" adage.

When to use one type of model vs the other?
Supervised learning models are great to use when we have a large data set of "known" outcomes that we would like to try to predict in the future. Unsupervised learning is excellent for exploratory use cases or unlocking previously unknown similarities between data features. The unsupervised learning models often follow a "What object am I most like?" approach in making its determinations.

Notice: Information contained herein is not and should not be construed as an offer, solicitation, or recommendation to buy or sell securities. The information has been obtained from sources we believe to be reliable; however no guarantee is made or implied with respect to its accuracy, timeliness, or completeness. Author does not own the any crypto currency discussed. The information and content are subject to change without notice. CryptoDataDownload and its affiliates do not provide investment, tax, legal or accounting advice.

This material has been prepared for informational purposes only and is the opinion of the author, and is not intended to provide, and should not be relied on for, investment, tax, legal, accounting advice. You should consult your own investment, tax, legal and accounting advisors before engaging in any transaction. All content published by CryptoDataDownload is not an endorsement whatsoever. CryptoDataDownload was not compensated to submit this article. Please also visit our Privacy policy; disclaimer; and terms and conditions page for further information.