Types of Machine Learning

1.    Supervised Machine Learning:

Supervised machine learning collects the data to train the model. In the supervised machine learning model, we require previous data and previous results as input. By training this data the result is predicted with good accuracy.

The collecting data must-have label. Suppose, in a bucket, there are some fruits, and if their names are specified it is said to be labeled. There are mainly two types of supervised machine learning problems.

       i.           Classification Problem

     ii.           Regression Problem

Supervised machine learning has various types of methods like Linear Regression, Logistic Regression, Na?ve Bayes theorem, Decision Tree, Random Forest, Support Vector Machine, Simple Neural Network, etc.

2.    Unsupervised Machine Learning:

Unsupervised machine learning has no labeled data, and the need for previous data as input is not required. It is self-organized learning and it finds unknown patterns in a dataset without labels.

The data has no label, but the algorithm helps to cluster the data by a similar category. For example, there is some data of cats and dogs, the model processes the data, and the algorithm divides the data into a similar category.

Unsupervised machine learning has various types of methods like K-means Clustering, Gaussian Mixture Model, Hierarchical Clustering, Recommender System, etc.

3.    Reinforcement Learning:

Reinforcement learning is like a reward system, it is an enforcing model to learn how to make a decision. Reinforcement learning follows the trial and error method.

There are two types of Reinforcement learning model:

       i.           Markov Decision Process

     ii.           Q learning

For example, a child trying to take the first steps;

Before he/she take the first step there are some instructions that he/she follows:

       i.           Observation

     ii.           Standstill

   iii.           Trying to balance the body

   iv.           Remaining still

For a child, it is a challenging task, but for us, it is very easy because we have learned it. If a child takes the first step, then he/she gets a reward, and if failed, then he/she doesn’t receive the reward. This is how reinforcement learning works.