Machine learning is a branch of Artificial Intelligence (AI) and computer science which focuses on the use of data and algorithms to imitate the way that humans learn, gradually improving its accuracy. — IBM
As you already know machine learning is a hot topic these days with the rise of Chat GPT and BARD.
Like human learns things with the help of other humans or reference books, computers can also learn using historical data.
If we elaborate on the definition of machine learning, it is a subset of artificial intelligence which develop algorithms that allow the computer to learn from historical data (training data) on its own. If it has a large set of data, it can improve the performance and can give more accurate predictions.
Common scenarios where machine learning will be used are image recognition, speech recognition, recommender system, and many more.
Machine learning can be classified into three types Supervised learning, Unsupervised learning, and Reinforcement learning. In this article, the primary focus will be Supervised learning and Unsupervised learning.
Supervised Learning
Supervised learning is a type of machine learning method that uses provided labeled data to the model to train it and predict the output. This is mainly based on supervision like a student learns the lesson under a teacher’s supervision.
Spam filtering is an example of an application in supervised learning. Human interaction is needed to accurately label data for supervised learning.
This type of machine learning is used to classify unseen data into categories and forecast trends using historical data. Usually, predictive models are trained with supervised learning by learning patterns between input and output data.
Supervised learning can be divided into two categories of algorithms as Classification and Regression.
Unsupervised Learning
Unsupervised learning is a type of machine learning method that learns without any supervision or human interaction. It is used to find trends in raw datasets or to cluster similar kinds of data into a specific number of groups (clustering).
The target of unsupervised learning is to restructure the input data into new features or a group of objects with similar trends and patterns. Like in supervised learning, in this type of machine learning, we don’t have a predetermined result. The majority of available data is unlabeled.
Unsupervised learning is most suited when answering questions about unseen trends and relationships within data. As an example, this machine learning type can be used to cluster customer data in marketing campaigns, to detect anomalies or outliers, and many more.
Unsupervised learning can be divided into two categories of algorithms as Clustering and Association.
Summary of the article
Supervised learning will learn the relationship between input and output through labeled training data whereas, unsupervised learning will find underlying patterns and relationships through unlabeled raw data by itself without any supervision as in supervised learning.