DEV Community

Panupog Chamsomboon
Panupog Chamsomboon

Posted on

Programming Rule Vs Machine Learning

Hi everyone, Today I'am seeing the people to activity such as post, comment, and upload something in social media, shopping with digital platform, control smart electrical appliance with application and other. This is reason to make a lot of data in every day.

A lot of data is incomming to make new problems such as Marketer cann't analyzing a lot of data is comming every day.
I'm web developer. I like to solve to problems. I would like to learn that how to solve problems in the news problem about a lot of data.

Now, Machine learning is one of solutions to solve problem about how to analyzing a big data. I'm starting to learn machine learning and start to learn with tensorflowjs becase I'm freindly to use javascript to make some services and applicaiton and I'm starting to read book Deep learning with Javascript. The book has overview the concept of programing rule and machine learning. This is good for understand and start learning machine learning

Image description

Programing rule is something you make application or some function. you write to rule when data input and return to something output when you write by rule.

Machine Learning is someting you have dataset and answer you putting to algorithm of machine to learn data and output to rule of data that cover the awnser when new data put the rule. The output of rule that we call Model.

Image description

The picture explain when you would like to use machine learning to solve problem in your application. we have 2 phase

Training Phase, First, You create machine learning model and need to have dataset and answer to trainning your model and when you trained model you have countinue to inference phase for using trained model

Inference Phase, Final, You apply trained model from model in Training Phase to your application or function in your code. when new data is comming to application or function the trainend model is processing your new data and output to your answer follow by tranined model that you make.

Discussion (0)