Overview of Classification

Hello,

I am mohit from LearnVern,( 6 seconds pause ; music )

In the Machine learning's previous tutorial, we understood about the types of Paradigms of Machine Learning.

In which, we learned about supervised learning, unsupervised learning and reinforcement learning.

And today, we are going to see Classification.

Classification comes under supervised Learning,

In supervised learning, we have understood that there is a supervision that is along with input data, we also have output, and we call them as labels.

So, let's see What exactly is Classification?

We will understand this by using our same previous slide example.

Here, we have taken a pomegranate, one lime, and mango.

You can see that the shape, color and size of lime are different and mango's shape, color and size are different.

So, here we should understand that we can identify any object by its unique color, size, and shape, and then can easily classify and identify them as mango, guava etc.

So, this is the way to classify.

Let's take another example,

Supposingly you have a 100 rupee note, and somebody else also gives you a 100 rupee note, then you can classify it as 100 rupee.

And if somebody gave you 10 rupees, would you classify it as 100?

No! You cannot because the shape and size are different, along with that the printed value on it is 10 rupee.

So, classification basically means to understand from the input and identify the class in which it's output should belong.

For instance, if you are walking under sunlight and someone asks you, is it day or night?

You can easily classify that this time belongs to day.

And on the other hand, if you are roaming under moonlight, then also you can easily classify the time being night.

So, in classification we define the classes before only and fit, adjust or predict these new inputs according to their classes in which they should belong.

Now, let's see this example ahead.

Now, looking at this shape and color you can tell as to which class it belongs, so this is an apple class.

And this belongs to the pomegranate class.

So, this is how classification works.

Understood ! Let's move ahead…

(02/48)

In classifications, we can have categories, for example red ball and blue ball, day and night win or loss, start and stop.

So, in all these examples, we will call them binary , Red one class and Blue one class, Day one class and night one class, that is, two classes or binary classes.

Okay !

Similarly, I have kept one more example for you, that is, money, holiday and movie.

So, if you have money and holiday both, then you will go for a movie.

If you don't have money and holiday both, then you will not go for a movie.

If you have money but no holiday, still you won't be able to go for a movie.

And if you do not have money but have a holiday, you still won't be able to go.

So, this is input one we call it as x1,this is input 2 also called as X2, and movie is output and here we can also call it as label, or target denoted as y.

So, this is also a binary class having yes or no.

So, let's see ahead

About multi-class classification.

In this more than two classes are present.

For instance, we had used an example of fruits like oranges, apples or pears, so multiple classes are present here.

So, similarly, if we have to find ranks as 1 St rank, second rank or third rank,

Here also we have more than 2 classes meaning multiple classes.

So this is multi class classification,

Now moving ahead,

We will just look at some names of algorithms, and learn about them later both conceptually and practically.

Here, first algorithm is support vector machine।

Second, naive Bayes classifier, k- nearest neighbours, random forest and decision tree.

So, these are some of the algorithms which help us with classification.

Their logic, implementation, and mathematical intuition are different from each other, which we will learn in detail in our upcoming sessions.

So, our next topic will be logistic regression, its name is regression but does the work of classification.

So, we will understand how logistic regression works in our next session.

Thank you for watching.

Remain motivated.

If you have any queries or comments, click the discussion button below the video and post there. This way, you will be able to connect to fellow learners and discuss the course. Also, Our Team will try to solve your query.