Tip: If you don’t have an Azure subscription, create a free account before you begin.

The last article discusses the basics of Machine Learning and Deep Learning, in this article, let’s leverage the Azure cloud service to implement the CNN to predict the malaria cells.

Step 1: After creating the Azure Free trial account, log in to Azure portal

Step 2: Search for Machine Learning in the search bar and click on create.


  1. What is Machine Learning?
  2. What is Deep Learning?
  3. What is Azure ML?
  4. Steps to Perform Image Classification in Azure ML (Python SDK).

What is Machine Learning?

Machine Learning is the application of Artificial Intelligence that focuses on the development of computer programs, through which it can constantly learn to perform tasks.

Click here to know more about Machine Learning.

What is Deep Learning?

Deep Learning is the subset of Machine Learning, which imitates the functionality of the human brain to learn and perform tasks.

Deep Learning is also called deep neural networks, and the most popular type of deep…

Every day, hundreds of thousands of customers give reviews on the products and services they buy in amazon. With customer opinion available on a scale like this, anyone can generate insights about their business, their competitors, and potential opportunities.

Already the reviews are scraped from the website and stored in a CSV format, Now let us generate some insights around the reviews.

Cleaning Reviews

This function helps us to clean the reviews, it removes all the stop words, punctuations, reducing a word to its word stem and the cleaned review is stored in the separate column.

Word Cloud

Decision Tree Learning

The decision tree starts with the root node consisting of the complete dataset and then uses an intelligent strategy to split the nodes.

classification tree uses information gain obtained from the entropy of the features to prioritize tree.


  1. Rules generated are simple and interpretable
  2. Do not require to normalize or create a dummy variable

Random Forest

A random forest is ensemble of decision trees, where each tree is built from samples with replacement and randomly selected.

The number of trees and the number of samples to be used in the random forest can affect accuracy.

Why do we need “kernel trick”?

When we realize that a linear boundary is not providing the better classification, this can be felt by poor accuracy or by some visualizations, after realizing that the going for a linear boundary will not work, we go for kernel trick

Kernel trick helps us to transform the variables into a higher dimension, where they are linearly separable. This helps us to classify the groups in a better way.

Importance of Data

Data must be readily available when it is needed. It is essential to run your business efficiently but is also critical to achieving compliance.

Web Scraping

How good it will be when we have data from the websites, which we are viewing through a web browser. There is no option of saving it anywhere.

Copying manually will be a tedious job, Web scraping is a technique which will help in automating the process of saving the required data from any website.

Setting Header and cookies

Python Requests does not force you to use headers while scraping but…

Campaign Effectiveness:

Running a successful campaign takes time, money, and effort. And after all of that, the question is: How to justify that our campaigns are effective?

Most of the people look at the sales figures to measure the campaign effectiveness. There’s a reason for focusing on sales figures because that's what creates revenue.

New sales are not the only sign of a campaign’s success. Different campaigns have different focuses. Certain campaign might not directly lead to an immediate increase in sales volume, it can still help us to grow our business through an increase in brand awareness.



Customer Lifetime Value

CLTV is the total value created by any given customer to the company. This helps Businesses to determine the acquisition cost for the new customers.

Stages of Customer Lifecycle


Reach creates awareness of the product. This invites the customer or the target audience to know and learn more about the product.


Reaching the potential customer won't help if you cant deliver the relevant product. contacting them directly with personalized communication will help in acquiring them.


Once the customer made his first purchase, the business starts to upsell or cross-sell its product to the customer…


Regression is an approach that gives certain weights or the coefficient for the various features which affect the target so that the error is minimum.

This technique is used to find the relationship between the variables. For Example the relationship between the advertisement expenses and the increase in sales

Here we fit a line to the data points so that the error is minimum.

Neural Network:

Linear Regression:

Linear regression is a approach which gives certain weights or the coefficient for the various features which affects the target, so that the error is minimum.

Linear regression helps us to forecast the future values of the target for the corresponding features.

Logistic regression

Logistic regression is the special type of regression which tells us probability of certain event.

Logistic regression is done on linear regression, which has a logistic function or the threshold value which decides where the event should be categorised.

Vijesh s

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store