Competition winners:

Pic Courtesy: Texas A&M Institute of Data Science Competition (2024)

Impact of corporate funding and ethnicity on the outcome of US Presidential elections

Texas A&M Institute of Data Science - Annual Competition: Winner (1st Prize)                                                                           Spring 2021

Team members: Sambandh Dhal, Tushar Pandey, Swarnabha Roy, Ritesh Singh Suhag

 

  • Devised a Return On Investment (ROI) metric to observe the effectiveness of party expenditure in each state
  • Analyzed the corporate funding & spending per state by categorizing them into different sectors over the 6 Presidential campaigns
  • Used t-distributed stochastic neighbor embedding to analyze the changing demographics of each state on the Presidential outcome in each election
  • Please see details: https://github.com/Sambandh/USElectionsAnalysis

Impact of Inter-Departmental collaboration on Research Outcome of Texas A&M University

Texas A&M Institute of Data Science - Annual Competition: Winner (2nd Prize)                                                                           Spring 2022

Team members: Sambandh Dhal, Tushar Pandey, Rohit Dube, Swarnabha Roy, Abhijit Mahapatra

  • Used NLP techniques like Unigram, Bigram and Trigram methods to determine the 30 most important topics within a department from publications
  • Devised Impact Score metric taking the normalized funding, number of publications & citations per department to calculate research impact
  • Visualized the amount and length of funding per department using Keppler Mapper and Network Analysis
  • Please see details: https://github.com/Sambandh/TAMIDS--2022

Improving Total Cost of Ownership (TCO) in a vehicle-driver ecosystem

Daimler US CESG Hackathon: Winner (3rd Prize)                                                                                                                                    Fall 2022

Team members: Sambandh Dhal, Shikhadri Mahanta, Krishna Chaitanya Gadepally

  • Analyzed the past GPS data on different routes at varying times of the day to gauge the traffic conditions, thus enabling the driver to maintain the engine health as well as optimize the fuel consumption

  • The real time GPS data from the trucks was used to monitor the traffic along different routes and update the
    routes in real time where 'Following Distance' was used as a metric

  • The historical active braking and lane departure sensor data was analyzed to gauge the efficiency of the current sensing system and give recommendations
  • The weather data was analyzed in real time along the entire designated route of the truck and if the conditions are
    adverse, the driver would be sent an alert in advance to avoid that pathway
  • Please see details: https://github.com/Sambandh/Daimler-CESG-Hackathon

Prediction and Inference of Large Wildfire Burn area in Contiguous United States

Texas A&M Institute of Data Science - Annual Competition: Winner (3rd Prize)                                                                            Spring 2023

Team members: Sambandh Dhal, Shubham Jain, Krishna Chaitanya Gadepally, Prathik Vijaykumar

  • Developed a predictive model using a 5-layered Deep Neural Network trained on the climatological attributes, drought severity index, land cover, and vegetative indices from satellite data sources to predict the spread of wildfires at 4,536 large wildfire locations in the US over the last 10 years
  • Conducted a SHAP analysis to visualize feature importance and partial dependence, which revealed the highest importance of land cover around the vicinity of wildfire occurrence for prediction of total wildfire burn area
  • Developed a mobile-friendly interactive web-tool to visualize the wildfire burn area and related datasets for the past decade
  • Please see details: https://github.com/Sambandh/2023-TAMIDS-Competition-Data-Sam-s-Strikers--main

 

Inference of Sea Level rise observed using Copernicus Data (Gulf of Mexico and East Coast)

Texas A&M Insititute of Data Science - Annual Competition: Winner (3rd Prize)                                                                           Spring 2024

Team members: Sambandh Dhal, Vivekvardhan Kesireddy, Tushar Pandey, Rishabh Singh, Sheelabhadra Dey

  • Analysis focused on sea level rise (SLR) data along the East Coast and Gulf of Mexico from 1993 to 2023.
  • Modeled greenhouse gas emissions, main contributors to global warming, using ensemble LSTM models to predict SLR in real-time.
  • Utilized a weighted model (SARIMAX, LSTM, Exponential Smoothing) to forecast SLR trend from 2024 to 2103.
  • Forecasted conductance and Dissolved Oxygen (DO) values influenced by SLR, crucial for marine life.
  • Validated approach through visualization of marine GDP data and cluster analysis of greenhouse gases contributing to global warming.
  • Recommendation: Transition to renewable energy sources, enhance energy efficiency to reduce greenhouse gas emissions.
  • Coastal protection and adaptation strategies are vital, including nature-based solutions like mangrove restoration and infrastructure investments like seawalls.
  • Please see details: Sambandh/TAMIDS2024 (github.com)