The Azure Machine Learning Guide: Part 1

DZone 's Guide to

The Azure Machine Learning Guide: Part 1

Rajat Jaiswal explains the fundamentals of machine learning in the first part of his guide to employing Azure Machine Learning in your project.

· Big Data Zone ·
Free Resource

Machine learning is not new but today it's a buzz word everywhere. You might realize that there are lots of things happening in the Machine Learning field.

Big companies like Microsoft, Oracle, IBM, SAP, and many others are working in this area. They have provided Azure Machine Learning, Oracle Advanced Analytics, IBM SPSS, and SAP Predictive Analysis astools to work with.

Before jumping into Azure Machine Learning, let’s first understand the basics of Machine Learning.

Machine learning is a way to understand the data pattern, recognize it, and predict accordingly for future.  It helps in:

1) Data Mining

2) Language Processing

3) Image Recognition

And many other Artificial Intelligence-related endeavors.

The above statement is bit bookish so let me explain in Indiandotnet style.

Let’s say you are a teacher in a school and you have a great deal of experience teaching. Each year, you teach many students and you also keep previous years data and basic details about the students.

Parents come to meet you and want to know about the progress and whether their child will pass and graduation or not. You simply do data analysis in your mind about whether that student studies or not, their percentage in the last couple of exams or internal assessments, how they performed in previous class, and then you give your prediction to the parents about whether their child will do good or bad in the final exam.

Now suppose that, instead of you, there is a computer and parents are asking the same question to the computer. A computer should provide the same answer as you give as accurately or even more so.

For this, we need to feed enough data in the computer, if the teacher has previous data samples by which the computer can analysis and predict accurately.

This overall exercise of processing data is part of Machine Learning.

So, firstly you have to train the computer by providing the initial data that we call training data. This is an iterative process.

Machine Learning, though, is more than this. Here are some more example where machine learning can help:

1) Detecting fraud credit card.

2) Determine SPAM emails.

3) Provide customer like to switch to competitor.

4) Free text when typing etc. many more examples of machine learning.

There are two distinctions in Machine Learning:

1) Supervised Machine Learning:

The supervised learning means the value you want to predict already exists in training data. It also means the data already exists in the computer, so the data is labeled. The accuracy is high in such a case.

2) Unsupervised Machine Learning:

This is opposite to Supervised Machine Learning. In this technique, the predictive data is not present in training data.

I hope now you have a basic understanding of Machine Learning. In next post, I will share a step by step example of Azure Machine Learning.

azure, big data, dotnet, machine learning

Published at DZone with permission of Rajat Jaiswal , DZone MVB. See the original article here.

Opinions expressed by DZone contributors are their own.

{{ parent.title || parent.header.title}}

{{ parent.tldr }}

{{ parent.urlSource.name }}