Here I will design a conjoint experiment and use Logistic Regression to estimate the impact of the features on consumer choice. The beautiful thing about this project is that I can collect the data myself through personal or internet surveys.
In this case, I will use a Bass Diffusion Model to forecast the adoption of a app on a new country, using data of the launch on a different country. While there are specific marketing related reasons why the app was download more often for a short period of time, the general pattern can be fit by a Bass Model.
Using an IBM Watson sample dataset, I will construct a Static Dashboard using R Markdown – instead of Watson – which is meant to represent the KPI reporting I used to do at Claro for the Fixed Line Internet Line of Business.
I have already used done something similar to this in another project, but this time I want to go deeper into the MMM methodology; specifically with respect to the lagged effect of advertising. There data for this comes from the book An Introduction to Statistical Learning.
I created this model in order to answer the question of which stores my previous employer should close. I basically defined the territory in terms of the telecommunication towers of the company, and iteratively decided where to open a store. The result is an optimal network of stores, and any store not in it should be closed.
Here I replicate a microeconometric analysis I did during my Master’s, where I used a Fixed-Effects model to identify the effect the share of Immigrant Students has on the academic performance of Native Students. I explain why this specific model was needed, and what is the rationale for why the effect exists.