This example is extracted from Lecture 4 notes from BAMA520 winter 2021.

General form of linear regression is listed below, where each term represents an ‘atomic’ impact of one x variable on the response.

In many cases, we require more complex model than that. Suppose factor A has some effect on Y, and factor B also has some other effect on Y. What if A and B work together? If their combined effort either augments or lessens the value of Y, then we say A and B has positive or negative interaction, respectively. For example, if drug A and drug…

Based on what we learned in BAMS517: Decision making under uncertainties , our group created an example to illustrate the concept of ‘ value of information’ . The example below is hypothetical and numbers given are guessed/random.

Suppose a scenario where a company is deciding whether to hold a running competition on a particular date in the summer , they could set the distance to be marathon, half marathon, or 1km . In addition to types, the profit of holding such a race largely depends on how the weather will be like on that day. Runners prefer cloudy days to…

ROC Curve , Cumulative lift chart, cumulative gain chart

I think it is appropriate to define and elaborate on the terminology ‘response model’ , to begin with. In typical situations on customer analysis, business analysts are concerned with building such a model. “The response model is a classification model. The task is to classify the customers who will respond to the next marketing campaign on the basis of information collected about the customers.”(source:https://www.sciencedirect.com/topics/mathematics/responsemodel#:~:text=The%20response%20model%20is%20a,objective%20variable%20to%20be%20modeled.) For a typical company with reasonable size and range of services, its customer base is large. In this huge customer base, there are different types of customers…