Machine Learning

Contents

• What is Machine Learning?
• Computer Programming vs Machine Learning
• Types of Machine Learning

"Machine learning is the automation of automation." - Jeff Hawkins

What is Machine Learning?

              Machine Learning (ML) is a type of Artificial Intelligence that allows computer the ability to learn without being explicitly programmed. The machine learning model learn from the past experience to predict the future outcome like humans. The historical data collected over the past years or months which are processed to fit into a model to find the unknown data outcome.

Computer Programming vs Machine Learning

              In programming, we know that the humans give instructions to machines to solve problems. We create a program and give data to the machine as an input to get the required output. But in Machine Learning, by providing data to the computer, the computer autonomously generates programs based on its computational capabilities and algorithmic intelligence.

Types of Machine Learning

-       Supervised learning

-       Unsupervised learning

-       Reinforcement learning

1. Supervised learning

              The supervised machine learning is a process of training the model with the data with known outcomes to predict the data with unknown outcomes. The features or variables or attributes are the independent variables whereas label or target is the dependent variable that is outcome. The supervised learning can be further divided into two problems, classification and regression.

2. Unsupervised learning

              The unsupervised learning is a machine learning technique that allows to discover hidden patterns or structures within data using ML algorithms. Simply we can say that in unsupervised machine learning, the data points have unknown outcomes which means the data set have only input features and no output variable i.e., target.

3. Reinforcement Learning

Reinforcement learning (RL) is a subfield of machine learning that focuses on developing algorithms and techniques to enable agents to learn and make decisions in an environment through trial and error. RL is inspired by the way humans and animals learn to interact with their surroundings.