What is Machine Learning?

Learning is the ability to take decision based on prior knowledge. Human learns things by experiences, which may come from reading, writing, listening or various other activities. If we have to define the human learning model it can be simplified as observing things, gaining knowledge and building our own models that help us to predict our behaviour in future.

In simple words, humans learn from their past experiences and use them to take future decision. In the same way the experience for the machine can be given as data which help to build analytical models to make future predictions or identify patterns.

In machine learning, an algorithm is trained on a data set which then uses the patterns it learned from the data to make predictions or decisions without being explicitly programmed to perform the task. For example, the algorithm can be trained by showing multiple pictures of cars and truck to recognize these object in a given picture.

Machine learning methods can be broadly categorised into four categories viz. supervised learning, unsupervised learning, semi-supervised learning, and reinforcement learning

Supervised learning is a type of machine learning where the algorithm is trained on labeled data, meaning that the data set includes both input data and the corresponding correct output. The algorithm learns pattern from it to make predictions for un-labeled data. For example, a supervised learning algorithm can learn from the characteristics of known patients and healthy individuals to build a classification model to diagnose diseased patients.

Unsupervised learning is a type of machine learning where the algorithm is not given any labeled training data. Instead, it must find patterns and relationships in the data on its own. One common application of unsupervised learning is clustering, where the algorithm groups similar data points together.

Semi-supervised learning is a type of machine learning that falls between supervised and unsupervised learning. In semi-supervised learning, the algorithm is given some labeled data and some unlabeled data. This can be useful when it is expensive or time-consuming to label a large data set, as the algorithm can still learn some useful information from the labeled data and apply it to the unlabeled data.

Reinforcement learning is a type of machine learning where an agent learns to interact with its environment in order to maximize a reward. The agent learns through trial and error, receiving positive reinforcement when it makes a good decision and negative reinforcement when it makes a bad decision. Reinforcement learning has been used to train machines to play games, control robots, and more.