Business Understanding for Machine Learning – Descriptive Analysis

Facebooktwittergoogle_plusredditpinterestlinkedintumblrmailFacebooktwittergoogle_plusredditpinterestlinkedintumblrmail

Business Understanding

Business understanding is the main and first step for doing machine learning in any platform or languages. Not all business problem can be addressed by machine learning approaches. There are some proposed categories for machine learning such as “Supervised Learning” and “Un-Supervised Learning”.

Supervised Learning is about when we identify both input and output variable while unsupervised learning is when we just have input.

Another machine learning classification has three major groups as

·        Descriptive analysis

·        Predictive analysis

·        Prescriptive analysis.

The descriptive analysis uses mainly unsupervised learning approaches for summarizing, classifying, extracting rules to answer what happens was happened in the past. While Predictive analysis is about machine learning approaches for the aim forecasting future data based on past data. And finally, prescriptive analysis use optimization and recommendation algorithms to answer possible outcomes.

However, there is other analytics like diagnosis analytics that also able to help people to get a better understanding of their business. According to the Gartner [1], these four analytics approaches help decision makers in the organization to make a better decision.

Each of this analysis (above figure) provides some values for the company and implementing them needs some skills level. Descriptive analysis is the easiest one that does not need a high level of skills and can be done with different tools. While diagnosis and discovery analysis able to provide more insight on data and needs more level of skills.

 Descriptive analysis

Descriptive analysis is the main approach to understand the data and existing patterns. In the Descriptive analysis can be used for understating the structure of the data, existing rules or patterns in data and so forth. Descriptive analysis can be categories into four main groups.
• Operational Reports
• Statistical Analysis
• Data mining approaches
Operation reports are mainly about the traditional reports that are about the analysis of the past data. These reports mainly use to judge the performance of companies.

Below figure shows a report that illustrates the average total hourly rate of labor based on different industries and income groups.

This report provides a brief description of the hourly rate of people in different industries.

The other reports related to some Statistical reports that show some statistical facts about data.

Statistical analysis is another way of doing descriptive analysis which able to help us a better understanding of the statistical behavior of data. There is various statistical analysis such as Univariate Analysis that we consider each column separately, such as a summary of data and data distribution. This report provides a brief description of the hourly rate of people in different industries. The other reports related to some Statistical reports that show some statistical facts about data. Statistical analysis is another way of doing descriptive analysis which able to help us a better understanding of the statistical behavior of data. There is various statistical analysis such as Univariate Analysis that we consider each column separately, such as a summary of data and data distribution. As you can see in the below picture we able to see the average hourly rate of the employee as a histogram to see how is the distribution, also we able to see the minimum, maximum, average. Median of this data in box plot or in a table format. The boxplot chart shows the data has been distributed smoothly as the median (the middle point) in the same as data average. 

The other statistical analysis is about the Bivariate correlation and regression between two different data column.

In the next Post, I will talk about the predictive and prescriptive analysis.

 

Facebooktwittergoogle_plusredditpinterestlinkedintumblrmailFacebooktwittergoogle_plusredditpinterestlinkedintumblrmail
rssyoutuberssyoutube
Leila Etaati
Dr. Leila Etaati is Principal Data Scientist, BI Consultant, and Speaker. She has over 10 years’ experience working with databases and software systems. She was involved in many large-scale projects for big sized companies. Leila has PhD of Information System department, University of Auckland, MS and BS in computer science. Leila is Microsoft Data Platform MVP.

2 Comments

Leave a Reply

Your email address will not be published. Required fields are marked *