Friday, May 27, 2022
HomeMachinelearningWhat is a machine learning model?

What is a machine learning model?

What is a machine learning model?

In the previous blog posts, we have learned about what is machine learning, how it is beneficial to human beings, and how to work with the CSV File which we called a Dataset.

So, whatever basic things we are required to know we have already got them.

Now let’s understand the Dataset in depth. Here I try to explain to you in very simple language because maybe you are totally fresher or have some experience. This Machine learning blog post series I have targeted for fresher as well as exp.

So, to understand from a fresher point of view, I can say Dataset is nothing but works as” training data”.

The algorithm will use training data and make decisions by itself. To create a well-managed ML code, There are many things to keep in mind.

  • What kind of result you want before creating an algorithm.
  • Create an algorithm that can make its own decisions.
  • Using Dataset and algorithm at the end we receive MODEL, which works as a backbone of ML.
  • (Next, question arises in your mind instead of Dataset can we use other thing? Well it is long topic so I will cover it later.)
  • All the test cases apply to the MODEL, like Whether it works as per the existing requirement or not.

For example (Let’s talk in a general term) your model works and analyzes only TEXT INPUT value and your test case contain some IMAGES then it will not work properly, not fit into our model. So at that time we refine our training data and re-train our algorithm.

It is the continuous process of training, testing, and checking that our MODEL is up to mark or not. Means our model fulfills our requirements or not.

How to use Model in machine learning?

Your Model can be used as it is or stored in the database, the File system in binary format, Create REST API.

From the beginning of this line, we talked about the MODEL, MODEL, MODEL. Then the question arises How to create a MODEL? Well,

We have data 😊

Using DATA, Perform data pre-processing. You can do this in several ways like import the dataset and necessary library, iterate, and find out the missing data, Encoding the categorical data, split your dataset, feature scaling, data cleansing.

If you are fresher then you know that In C Programming all the code is related to O and 1.

The same way works here, at the time of data pre-processing it’s a good habit to convert your data in digit form like, for example, YES is equal to 1 and No is equal to 0. Because the model works with digital form data.

After performing all the above steps we receive our Model.

But make sure that you perform test cases again and again for better results because this is not the final MODEL. The model needs to keep itself updated at all times. In the end publish your Model in version, which can be used in various applications.

What is Supervised and Unsupervised Model?

In this blog post, we will get an overview of supervised.

Supervised, again divided into different categories like Classification, Regression. Here the algorithm learns from the data/features. We tell it what to look for.

Classification: –

  • The most prevailing application of Machine Learning.
  • The classification problem focuses on literally finding the classes to which the data belongs. That means the prediction of class in classification based on the input number/data.
  • In our previous blog post, we have discussed employee Salary and Related state. If you haven’t seen that post click here.
  • Based on the salary as well as from which state, we can identify that the employee earns how much amount monthly and based on that either he will buy the product or not. Again here we use boolean values like 1 for buy and 0 for not buy.
  • Here it will check the Matrix of feature (Click here to read more) and generate the model.

Regression:-

Regression does not determine the class, it predicts the continuous value. In an upcoming blog post, I will provide you with an example of Regression.

Here comes the end of this blog post, if you like my blog post please provide me your comment, feedback and share it on social media.

RELATED ARTICLES

LEAVE A REPLY

Please enter your comment!
Please enter your name here

- Advertisment -

Most Popular

Recent Comments