Types of machine learning | supervised, unsupervised, semi-supervised and reinforcement

 Lesson - 2

supervised, unsupervised, semi-supervised and reinforcement learning

TYPES OF MACHINE LEARNING

So as mention above we basically have four types of classification in machine learning algorithms you probably thinking that what the hack was these types so let's understand them one by one...

Supervised learning

def:: "supervised learning is when the model gets trained on the labeled dataset".

More simply we can say that in supervised learning the computer is presented with input data and their desired output data as well.
now, you may ask what that labeled data set actually is AAA okay, a label is nothing but a tag like an apple, cat, table, etc.

example 1- suppose we want to sum two numbers so instead oftelling the machine that the formula for adding two numbers is a + b. insted we give it some input data like if a=2 ,b=3 and the output is 5 and if a=4,b=5 then output is 9  so now if we ask the machine what is the output of a=2,b=5 then it automatically gives us a output based on what the machine learns from the above two sample data. if it is the desired output that is 7 then the given examples are enough but what if the answer is some other number then 7 so now its time to provide more input data to machine.

example 2- suppose you are given a basket filled with different kinds of fruits. Now the first step is to train the machine with all different fruits one by one like this:

   If shape of object is rounded and depression at top having color Red then it will be labeled as –Apple.

  If shape of object is long curving cylinder having color Yellow then it will be labeled as –Banana.

Now suppose after training the model(machine), you have given a new separate fruit say Banana from basket and asked to identify it.


Since the machine has already learned the things from previous data and this time have to use it wisely. It will first classify the fruit with its shape and color and would confirm the fruit name as BANANA and put it in Banana category. Thus the machine learns the things from training data(basket containing fruits) and then apply the knowledge to test data(new fruit).

I guess now you get a fair idea what this supervise learning thing is OK buddies its time to explorer the subparts of supervised learning lets understand them too...

Supervised learning is classified into two categories of algorithms:

1.Classification:

"A classification model predicts discrete values"                              OR                          

More easily we can say that a classification problem is when the output variable is a category, such as “Red” or “blue” or “disease” and “no disease”.                                           

For example, classification models make predictions that answer questions like the following:                

➤ Is a given email message spam or not spam?          

➤ Is this an image of a dog, a cat, or a hamster?                                                                                                                                               

2.Regression:

  "A regression model predicts continuous values"                   OR                             

A Regression problem is when the output variable is a real value, such as “dollars” or “weight”.

For example, regression models make predictions that answer questions like the following:

➤ What is the value of a house in California?

➤ What is the probability that a user will click on this ad?

okay, now I am pretty much sure that your doubts about supervise learning get resolved if not you can always ask in the comment section.

ooh I got this you do not get the full understanding of regression and classification plus the diagram provided makes the situation even worse. 
don't worry about those confusing diagrams we'll cover them in detail in a separate article.

let's move on to the next section that is unsupervised learning.

Unsupervised learning:

def:: "unsupervised learning is when no labels are given to the learning algorithm, leaving it on its own to find structure in its input".

pretty difficult huh.. let's make it simple okay so unlike supervised learning,  in unsupervised learning there is no prior training provided to the algorithm, therefore, the machine is restricted to find hidden patterns in the unlabeled data.
this means here we don't tell the machine that this is an apple or banana
the machine should figure it out on its own. 

example 1: suppose we have an image having cats and dogs both but the machine never seen it before now the twist here is that in supervise learning we train the algorithm how cat is different from dog but in unsupervised learning we do not give the algorithm any prior training  in addition ask the algorithm to classify them based on its own understanding

thus the machine have no idea about the features of dogs and cats so it can't categories them with labels like dogs and cats but but but it can categories them based on their similarity's and differences like dogs have long tongue cats have small, dogs are longer cats are shorter and you may make many more differences but how machines do that we'll get back to its technical details in our next articles

it's interesting to know how supervise and unsupervised learning are completely opposite brothers.
OK let's not waste any more time and explore the roots of machine learnings second brother that is unsupervised learning.

now here in unsupervised learning we again have two categories:

1.  clustering:
   
   
   
   clustering means making different groups as we know we can't classify objects in unsupervised learning but to do so we can make clusters(groups) of the same feature objects.                                                                                                                                                                                  
2. Association: in association the machine tries to relate two or more things based on some rules, such as people that buy X also tend to buy Y live example is YouTube for example when you watch the funny video it recommends you other funny videos.

so far we have covered the two major topics of machine learning now its time to cover the two remaining one that are-

Semi-supervised learning

 when a large amount of data is given in which some data is labeled and some data is not labeled then it is called semi-supervised learning so we can say it is the half half version of supervised and unsupervised learning.

example: suppose you were given a basket of fruits some have labels like apple,banana etc but some are unlabeled  so semi-supervised learning basically deals with such type of data.

Finally, we have our last topic lets talk about it :

Reinforcement learning

reinforcement learning is the real challenge in machine learning here the machine deal with the dynamic environment that is in the game of chess machines needs to take decision based on opponents moves or machine needs to play a shooting game like pubg on its on without human help so this is called reinforcement learning in this we try to implement our all understanding of machine learning.

don't worry we'll end this course with implementation of all our learning.

BOOKS FOR MACHINE LEARNING:

these are the books which I personally prefer hope you like the way of teaching too...

• A Guide To Data science

• Introduction To Machine Learning

Youtube Channel:

need the support of you guys - link to youtube

---

That's all for today hope you're liking my efforts keep enjoying and keep learning 

😃😃😃

---