Overview

Input is given to a machine learning model which then gives output according to the algorithm applied.

If the output is correct, we take the output as a final result else we provide feedback to the model and ask it to predict until it learns.

Basic Example

Paul is listening to music, can we predict if he will like a new the song?

Song A - This song clearly falls into the grouping of songs which Paul will like.

Song B - At first, it is not clear if Paul will like Song B since it lies between his grouping of like/disliked songs. Using K-Nearest Neighbours algorithm, we can assume that Paul will like Song B.

More Data → Better Model → Higher Accuracy

Types of Learning

Supervised Learning (Labeled Data)

Here the machine knows the labels associated with some features. Based on the features, it can predict the lablel.

Unsupervised Learning

Reinforcement Learning

• Reward based learning

Feedback to the model helps it with future predictions.

Applications of Machine Learning

References