top of page

Academic Projects

Graduate (MS in Business Analytics)

  • Predict Fuel Flow in Airplanes:       (Link to Presentation)

    • Identified key drivers of fuel-flow in different phases of flight

    • Major impact on fuel-flow by 'Control Column Position' during take-off

    • Impact on 30% operating costs, if optimised

 

  • Predict Loyal Customers:                (Link to Presentation)

    • Used data worth 100K rows to predict the loyalty propensity of a customer prior to initial purchase

    • Estimated savings up to $44K upon selective targeting

 

  • Predicted World Series Champion:

    • Data cleansing by removing redundant variable and KNN missing value imputation

    • Used logistic regression to predict 2016 's World Series Champion for BaseBall

    • Used Scikit-module in python

 

  • Predicted Car's Prices

    • Explored multinomial linear regression, random forest, gradient boosting and ensemble of gradient boosting with random forests

    • Got the second best result in class on test data 

    • Used R as a programming tool

 

  • Cross-Sell Opportunity(Capstone) :        (Link to Presentation)

    • Used a combination of Factor analysis, cluster analysis and product pairing propensity to identify top 20% targets

    • Estimated increased revenue of 470MM 

Undergraduate (B.Tech in Electronics and Communication Engineering)

  • AZTEC (Amplitude Zone Time Epoch Coding)

    • Explored usage of AZTEC for noise reduction

    • Analysed hundreds of noise signals for low error compression-decompression algorithm

 

  • Micromouse

    • Built a micro mouse, a robotic mouse that can solve complex maze

    • Used ARM Coding for PIC microcontroller to program the barrier handling algorithm

bottom of page