TYPES OF MACHINE LEARNING ALGORITHMS

So, I am back with Different Types Of Machine Learning Algorithms, and explained supervised and Unsupervised Learning in my previous Article , if you haven't seen that , do have a look .

So, various types of Machine Learning Algorithm :

Supervised Learning

So, if you are training your machine for every input with corresponding output(target), it is then called as supervised learning which will be able to provide target for any new input after sufficient training.

1. Classification :

Classification means the the action or process of Classifying/Categorizing something. Here , it means we will divide the data into classes , and will predict the categorical response.. In other words we are trying to use data to make a prediction about a discrete set of values or categorizes.

When the data can be classified into two classes ,it is called as Binary Classification.

 An example of a classification problem could be analyzing a image to determine if it contains a dog or a not.

Analyzing medical data to determine if a certain person is in a high risk group for a certain disease or not.

Will there be an event taking place or not .

Is this mail Spam or Not.

So , these are various Binary Classifiers , that is either YES OR NO (0 and 1 type). And all kinds of this problem can be solved using this approach.

When the data is divided into various different classes , it is called multi-class classification.

Analyzing the data to find out , whether it is a Dog , Cat , Fish , Tiger , Rabbit .

Is the an Apple , Orange , Banana .

So , when we divide our data into more than two classes .

2.Regression:

This is a type of problem where we need to predict the continuous-response value . Few Examples are :

what will be the price of house in a particular specific city??

what will be the salary of a person based on its experience ??

what will be price of stock ?

So , these kind of problem comes under Regression .

So , i hope we have got a basic idea of the types of supervised Learning

And if you are training your machine , only with a set of inputs, it is called unsupervised learning , which will be able to figure out the structure between different inputs. And then by using the concept of clustering and dimensional Reduction , it predict the output for the new inputs values .In today's world, Clustering is widely used .

Clustering : In this we divide our data into clusters , people generally get confused in multi-class Classification and Clustering , but there a key difference , in order to figure out when to perform clustering and when not to, depends on the data Sets , if its unlabeled we will perform clustering else can go with multi-class Classification.

Example :

If I invite you to a party where you meet totally strangers. Now you have no prior knowledge and this classification can be on the basis of gender, age group, gender or whatever way you wish to. So , here we have no prior Knowledge , that is why we will use unsupervised Learning instead of supervised.

Now , one of the most important type is the REINFORCEMENT TYPE

REINFORCEMENT LEARNING :

In this , it learns by interacting with the environment , and in this we have a system of punishment and reward and it learns from it , to give an output .

All dear readers ,i will love to explain it with a story of an Ant , the ant use to climb a place to get food for it . It used to fall down again and again but in order to get a food for her , it again and again climbs . It learned from its environment and every failure how height can be achieved.

So , we are over with basic understanding and friends , i want to know that are you all familiar with python or not , if you are not we will first start with basic knowledge of python , so comment Down and let me know .

#Learnt a lot

#Easy way to learn Machine Learning

Pallavi Goel