- Blogs
- Afia Ahmad's blog
- How can we Predict Outcome Events?
Home > Blogs > How can we Predict Outcome Events?
Plenty of business users are interested in being able to make predictions. At their simplest, predictions are made to predict variables such as the amount of sales to be expected for a given advertising spend, or other such questions. The next level of predictions are more complex - trying to predict the sales at a store, given that the sales depend on the store size, average number of walk-ins, number of SKUs the store carries, average pedestrian and vehicular traffic in front of the store.
The difference between the two examples lie in the difference in the number of independent variables, or predictors. Predictors could be one, or many; and when many, the most important predictors (the ones with the highest predictive ability) can also be identified. The similarity of both the above examples is that the outcome is a continuous variable. The predicted sales could, for example, be INR 90,02, 379. Or we could predict it to be, with reasonable certainty, anywhere between INR 78,32,293 and INR 1,10,23,992.
However, not every prediction in life lends itself to such answers. I sometimes ask Google Now on my phone, “Will it rain here tomorrow?” I’m not expecting a point estimation as an answer to my question when I ask this. It’ll be strange to get an answer as “3.2 mm”, giving an actual estimate of the rain. Similarly, saying “0 mm” is also not ideal. The preferred responses to the question would simply be a “yes” or a “no”.
Similarly, in business, a bank may want an answer as “yes” or “no” to a question about the chances that a potential borrower would default on repayment. A life insurer may want to base the premium on the chances that the insured may die. Clearly, the regression that we usually think of, would not work in these cases.
What can we do to answer these questions?
Answers in the next blog.
>> To know more about Business Analytics with R, or Business Analytics please go to Manipal ProLearn.
Soumyadip Pal is a retail analytics professional and a passionate educator with more than 8 years in the industry and more than 7 years in the academia, currently working as a consultant with Manipal Prolearn.