Research Publications

• Deep Reinforcement Learning for Autonomous Robot Navigation in Dynamic Environments
  Institute of Electrical and Electronics Engineers (IEEE) – Link

  ▪ Built a DRL model with TRPO, increasing accuracy 30%, cutting overhead 20%, and improving efficiency 25%
  ▪ Developed a CNN-based vision system, boosting detection 27%, reducing false positives 32%, and speeding processing 18%

• Machine Learning based Big Data applications for E-Commerce Logistics
  International Journal of Engineering, Science, Technology and Innovation (IJESTI) – Link

  ▪ Built a machine learning–based customer segmentation model using CNNs and gradient boosting, boosting targeted marketing efficiency by 20% and increasing purchase conversions by 15%
  ▪ Mapped and analyzed supply chain process data (order cycle, inventory turnover) to define performance metrics and improve end-to-end visibility using automated data pipelines

• Rides Request Demand Forecast – OLA Bike
  International Research Journal of Engineering and Technology (IRJET) – Link

  ▪ Applied clustering to segment ride zones and modeled demand-supply gaps, contributing to process improvements in real-time ride allocation and fleet optimization
  ▪ Optimized real-time ride allocation by 22% using Mini-Batch K-Means, clustering 4M data points into 50 zones

• Hybrid Reinforcement Learning based GA for Portfolio Optimization
  Journal of Northeastern University – Link

  ▪ Developed a Q-learning GA for portfolio optimization, increasing returns by 22%, reducing risk by 18%, and cutting error by 49.7%
  ▪ Enhanced investment strategy efficiency with 10% higher stability, 27% faster convergence, and 15% lower computational costs