QM901x – Predictive Analytics Prof. U Dinesh Kumar Module 1 – Clip 01 Week 3 Introduction to Analytics Hello Welcome to the Course Predictive Analytics. My Name is Dinesh Kumar and I am a Professor at the Indian Institute of Management Bangalore. I will be the instructor for the course. There is a lot of Buzz around Analytics. I am going to discuss why Analytics is very Popular and How companies are benefiting by use of Analytics. Today Many beliefs that the next economic growth will be driven by Analytics and Artificial Intelligence. Let me First Introduce the Business Analytics First. I Would like to start with the famous quote by Edwards Deming, he said "In God we trust, all others must bring data". He probably mentioned this quote in 1930's or 40's, long before terms such as data scientist or analytics or Artificial Intelligence was in use. Let us try to understand what may have triggered Deming to come up with this particular quote. When we talk about decision making, many organizations actually use different approaches. What happens is that, when organizations face problem they form a committee. And one person in the committee will immediately form a Whatsapp group and send irrelevant things. In the committee meetings, members of the committee express their opinion about what decision to take and how to solve the problem. There is one famous algorithm called "Hippo Algorithm" which stands for "Highest Paid Person's Opinion". Typically in most cases they may go with the opinion of the highest paid person, usually it may be the Opinion of CEO or CFO. When organization take decision based on highest paid persons opinion, they can go wrong. So that's what probably triggered the quote Deming, “In god we trust all others must bring data. The Quote emphasizes the importance of data and analytics in Decision Making. Let us try to understand what is Business Analytics? © All Rights Reserved, Indian Institute of Management Bangalore QM901x – Predictive Analytics Prof. U Dinesh Kumar Module 1 – Clip 01 Business Analytics is a multidisciplinary field that uses expertise such as statistical learning, Week 3 machine learning, artificial intelligence, computer science, information technology and management strategies to generate value from data. It has three main components Business Context, Science and Technology. Business Context is important since the success of Analytics will Depend on the ability to ask rights Questions. One of the Frequent Examples quoted by many about successful application of analytics is Target’s Pregnancy Test. Target is one of the Largest Retail Chains in India and early 2000, they developed a analytical model to predict whether a customer is pregnant or not. You may wonder why target was interested in Knowing whether someone is Pregnant? Well, they are special customers since they are likely to price insensitive. They would like to buy the best product as long as they can afford it during pregnancy and immediately after the child birth. It was estimated that a typical American parent spend about 7000 dollars before the child reaches its first birth day and the Market Size of Mom and Baby Products was close to 37 Billion USD. So, We are Talking about hugh market of Price insensitive customers, and capturing them early will add significant value. The Next Component of Business Analytics is Technology. Taking the Target’s Example, they have to collect transactional data of customer purchases, store it, retrieve it and analyses it to gain insights. This will require Data Base Systems, Software Tools to analyse the Data etc. In case of Big Data, which I will discuss later, we will need sophisticated technologies to handle huge volume of data. The Third Component of Analytics is the Science Part. Taking the Target’s Pregnancy Test Example, basically we have to classify a customer as either Pregnant or not Pregnant. In Analytics this is known as classification problem. Problems such as customer churn, employee attrition and fraud are examples of classification Problems. There are many algorithms used for solving Classification Problems such as logistic regression, decision trees, random forest, neural network and so on. The role of Data Science component of analytics is to find the best algorithm for a given problem based on a selection criteria. © All Rights Reserved, Indian Institute of Management Bangalore