Simplifying Analytics from Day-to-Day Events

Rohan, Charan and Sanju are doing MBA from a city college (not an A grade college). While Rohan is above average in Mathematics, having done his engineering from a private engineering college. Charan and Sanju are scared to their bones of this subject — MATHS. Analytics emerging as one of the important skill required in the industries and getting a lot of news of use of the same in several big companies with high package makes them drool over this subject, but the problem lies in understanding the subject. Rohan can somehow grasp something related to this subject and can improve if given proper guidance and more positive exposure on the subject, but the other two friends count sheep during the lecture and have almost given hope only to dream of working in traditional roles in companies.

What wrong is going on in the life of Charan and Sanju? Did they commit a crime by not knowing about Mathematics/Statistics? Maybe not, at least I do not think so, I feel that they lost interest and gave up, as they were not able to understand the basics because of fear of Mathematics and Statistics. They are the same people who ask that they have never seen the use of ( a + b ) 2 in real life. Well at least I have found a use of the same, it helps me to get the value of square of 17 apart from mugging it up–

Well many will laugh at this example, but I find this useful, as I cannot memorize things for a very long time so it may take me some seconds more to find out the value of square of 17, but I get it. Next time when someone asks you the use of such mathematical formulae, you can reply to them that they are used in life during adverse conditions and maybe you have not put yourself in a difficult situation in life, as you have not taken risk so you are asking me such questions. Jokes apart, let us take some daily life event and try to understand the application of analytics and then take big reference to its business applications.

Before going to the same, let us be more familiar with what is an equation; let us not follow the bookish definition. An equation is a mathematical representation of outcome of any phenomena or any event. Let us dig into it in a way that is more childish, many have study in their school what a Pythagoras theorem is:

To find the diagonal length of your room, you need to fit the length of two sides of your room in the given formula. So, if you know the length of two adjacent sides of your room, you can find the diagonal length of your room. The dimension of screen of TV, Laptop, Mobile etc. in different websites are diagonal lengths, so you know what you can do with the lengths.

Many know about the probability theory of tossing the coin. Tossing the coin is the event and its outcome is represented by a probability formula or an equation of probability –

Probability of Head = no of outcome of head / no of total outcomes

It can be seen that how an event can be represented in mathematical form to predict it’s results. Let’s not complicate the matter by thinking how we got the formula as deriving the formula is completely different part and let’s leave it to Mathematicians and experts in that field and concentrate on how to bring into use their findings into real life, or many analytics experts will call it model.

Let us take Simple regression and blend its application with some of our life event and decision-making. Before proceeding ahead, lets’ try to answer it in a simple way why Regression is used? Its use is to make a model or formula for trying to predict an outcome because of several reasons affecting the outcome, just as the outcome of tossing a coin is affected by presence of two sides of the coin, or the length of the diagonal of the right-angled triangle is dependent on the length of the other two sides. In simple (binomial or multinomial) regression, we can establish the relation between the outcome and the dependent variables affecting them by making a mathematical equation, where all these things are imbedded.

The equation, which you must have seen, is most commonly in this form –

We know that a health person’s body weight is impacted by the amount of food intake, hours of exercise they do, amount of water they consume, hours of sleep etc. What will you think if I say that a simple life event like measuring your body weight with variation of hours of exercise, food intake, water intake, hours of sleep and more, can be represented in the form of mathematical equation? Let us not get confused but align this body weight with previous formulas of Pythagoras theorem and Probability of tossing a coin. All the equations have left hand side representing the outcome or result or what we want to know which are what will be the outcome after tossing the coin, what will be the length of the diagonal of that triangle, and in our final regression case, it is body weight, which is represented by Y. The right hand side contains the value of all the variables, which directly affects the outcome represented by X1, X2… like length of two adjacent sides of the triangle, two sides of the coin and in the last case they are hours of exercise, food intake, water intake, hours of sleep, and there can be many more, the more the no of variables you include, the longer the equation will become and chances of more accurate output is high, however may become more complicated. You must be asking in your mind what the C, a, b, c … are, C is called he constant and a, b, c… are called co-efficient, which we get after processing the data in statistical tools which is describe ahead.

So how to do that? You need to measure and note down your body weight daily for certain period — say 6 months along with other variables’ value which are hours of exercise, food intake, water intake, hours of sleep. Now you can use these data collected in any statistical tool like SPSS or programing tool like python, R, but the simplest way is use of Microsoft Excel. There are a lot of blogs and videos available over internet where you can learn how to use these tools for regression, you just need to follow the methods shown and do not need much of your brain to exercise for that. These tools are like machines where you feed them your data and they give you the output as these mathematical equation or models, if operated properly. After running the process you will find a simple relation of the weight of your body with other activities that directly affects your weight in mathematical form as shown in the last equation Y = C + a*X1….. And you get the values of C, a, b, c…

The most interesting thing you will find that if you have been able to establish the equation correctly, then you can predict your weight with variation in those dependable activities considered. You can see the fun here and can use this equation or model to control your weight in future. If you will learn Logistic regression then you can make a model which will help to understand your own buying decision and you can precisely understand which factor influences you more to buy any product. You can predict many such things from your life if you have past data of Y and X1, X2 i.e. you know what to find and what are the things affecting that requirement.

Now how does a manager bring in the use of these in their daily work in an organization?

· A sales manager can build a regression equation to predict the sales of his product with several variables which affect the sale of the product for example sales of soap is affected by price, competitors price, discount, free offer, availability etc. now you know that Y or the output is Sales and other things described are the dependent variables. After building the model, he can predict the sales in near future using the mathematical equation. There are many such techniques and theories which can be used to develop model to predict if their product will be Bought or Not, can determine what price should be quoted, what should be the product feature favored by their target segment, etc.

· Any Investment firm can use analytics to build such model to predict the movement in the share market and take decision to invest in any firm or not, Different financial models help you to take financial decisions in a more logical path.

· HR manager can build a model to know training requirement based on models build to know productivity of employees etc. whether they will join their organization or reject the offer, similarly a model to predict retention and many more.

· Even in important big auctions, decisions of buying or leaving at a value is guided by analytics, remember the auctions of players in big leagues? Also many players, teams, coaches are using analytical tools to improve the game.

· You must have noticed how some companies feed your search on internet with several advertisement based on your past searches and how some companies feed your screen with an add with exactly your requirement? How prices of some services and products differ from time to time on their web application? Do you smell a big role of analytics in this?

Tools like Excel and SPSS can be used when no of data is small, but when you are dealing with huge no of data, you need to use more advanced tools like Python, R, SQL etc. Now think that these data are collected and manipulated by computer using these analytics tool, here comes the Machine learning and Artificial Intelligence. You are asking your computer or machine to give you with solution with the collected set of data. This is how world is becoming.

The explanations and examples might have helped a bit to reduce the fear about mathematics, statistics, numbers and big formulae. If it has really helped you to get a rough idea of what can be done in analytics then don’t stop to explore more about this field and start with statistics. If even an iota of fear has been reduced in you after reading this, then trust me, you can overcome a lot of hurdles in this subject. You just need to shun hesitation and keep exploring, information in just at the tip of your figures in this world now. Next time when you start to think on any outcome of any event then give more time to think about the factors/variables affecting them too, and you get more analytical in your approach to decision making.

If you still doubt, then start with this — You need to find the output on your left hand through the input variables in your right hand.