This repository contains all my personal projects and mini projects/assignments that are part of curicullum at UT Austin. As there are multiple projects in this repository, i have created readme documents for each project separately. Here you can find the introduction to each of those projects. For more detail, please look into the respective project folder
Association rule mining is a very interesting and important topic in retail analytics.Being able to find the associations between various products or services is the first step towards providing personalized recommendations.In this problem, i have tried my hand at finding the association rules between the different grocery items in the cart and visualizing the result in a user-friendly way.
Author attribution is one of the interesting applications of text analytics. In this exercise, i have tried to attribute the articles to various authors using different text pre-processing techniques and compared the performance of various classification models on the final dataset.
Market Segmentation has always been the first step in any product launch, campaign or personalized recommendation. This was data collected in the course of a market-research study using followers of the Twitter account of a large consumer brand that shall remain nameless---let's call it "NutrientH20" just to have a label. The goal here was for NutrientH20 to understand its social-media audience a little bit better, so that it could hone its messaging a little more sharply.
Predicting the test time for Mercedes benz cars after manufacturing was the main objective of this project. Using XGBoost, we achieved an R squared value of ~55% for the prediction model.
Understanding how customers are associating a brand with their own sentiments is crucial information to growth in industries. In this mini project, we have found associations between luxury cars discussed in Edmund’s forum and generated insights regarding what attributes are customers talking about when it comes to these brands