Skip to main content

web page@ MINTU YADAV






Mintu Yadav - IEOR, IIT Bombay


Industrial Engineering and Operations Research

Indian Institute of Technology Bombay

Personal Details

M.Tech (2023-2025)

Roll No: 23M1529

Gender: Male

Academic Background

Examination University Institute Year CPI/%
Post Graduation IIT Bombay IIT Bombay 2023-2025 8.55
Graduation Barkatullah University BUIT 2015-2019 7.5
Intermediate/+2 D.A.V Public School D.A.V Public School 2013-2015 8.8

M.Tech Research

  • Medical Prescription Generation (Master's Thesis) | Prof. P. Balamurugan, IIT Bombay
    • Developed a custom NER model to extract Proteinuria, Creatinine, GFR, and Age from RCT literature on IgA Nephropathy.
    • Fine-tuned BioBERT and BART with Iterative, Direct, and Bootstrap Fine-Tuning, achieving 0.995 and 0.953 accuracy, respectively.
    • Implemented a research paper ranking system using a scoring algorithm to retrieve relevant studies based on patient parameters.
    • Evaluated search effectiveness using NDCG metric, achieving 0.9789, outperforming Google (0.9116) and PubMed (0.9322).
  • Modified Particle Swarm Optimization (Seminar) | Prof. Jayendran Venkateswaran, IIT Bombay
    • Developed MPSO optimization algorithm, achieving up to 42% faster convergence compared to standard PSO technique.
    • Tested on 14 benchmark functions, showing up to 35% accuracy improvement in 11/14 evaluations on various datasets.
    • Conducted statistical analysis, confirming the significance of improvements with a p-value 0.05 using hypothesis testing.

Technical Projects

  • Dynamic Pricing for Ride-Sharing Services
    • Developed a dynamic pricing strategy that adjusts ride costs based on real-time demand and supply levels in data.
    • Implemented a pricing model with demand and supply multipliers to adjust prices for sudden demand and supply fluctuations.
    • Designed a dynamic web platform to predict ride-sharing prices using a well-trained Random Forest regression model.
  • Facial Recognition Attendance System
    • Developed a facial recognition-based attendance system using OpenCV and KNN, fine-tuning ResNet to achieve 95% accuracy.
    • Implemented real-time attendance tracking with liveness detection and instant verification, handling up to 100 images/user.
    • Designed and integrated a scalable web tool for attendance visualization with secure access control and exportable reports.
  • Mortgage Default Prediction
    • Conducted exploratory data analysis and feature selection using Extra Tree Classifier to extract top 10 crucial features.
    • Applied SMOTE for oversampling, utilized PCA for dimensionality reduction, and compared LR, SVR, and XGBoost Regressor.
    • Achieved AUC 0.83 and F1-Score 0.87 by XGBoost Regressor, significantly elevating risk management for banks and investors.
  • Automated Financial Dashboard with SQL and Power BI
    • Analyzed data using SQL queries and Power Query, seamlessly integrated SQL Server for advanced data transformations.
    • Improved ETL processes by controlling SQL Server jobs, enabling smooth Excel integration and real-time synchronization.
    • Enhanced financial dashboard functionality, streamlining data analysis and reporting for improved business insights.
  • Passenger Traffic Forecasting for Airlines
    • Applied Box-Cox transformation, conducted ADF and KPSS tests to verify dataset stationarity and consistency.
    • Leveraged ARIMA, SARIMA, Holt-Winters, and ARMA methods, uncovering intricate complex patterns and seasonal trends.
    • Achieved superior predictive accuracy with SARIMA, yielding metrics of 38.00 RMSE and 8.00 MAPE, ensuring reliability.
  • Mumbai Airport Runway Simulation
    • Simulated and optimized runway operations using AnyLogic, achieving 92.69% utilization and reducing waiting times.
  • Undirected TSP Problem
    • Solved the Traveling Salesman Problem using Pyomo for 25 cities in Maharashtra, optimizing routes with Integer Programming.
  • Transfer Learning for Survival Analysis
    • Implemented a transfer-Cox algorithm for cancer survival analysis, achieving a C-index of 0.5869.

Technical Skills and Key Courses

  • Key Courses:
    • Machine Learning
    • Engineering Statistics
    • Deep Learning
    • Mathematical Optimization Techniques
    • Probability and Stochastic Processes
    • Multivariate Data Analysis
  • Languages and Tools: Python, SQL, Power BI
  • Libraries: Pandas, NumPy, Matplotlib, PyTorch, Transformers, LangChain, OpenCV
  • Additional Skills: PySpark, Excel, Git, Google Analytics, Simulation (AnyLogic), Optimization (AMPL, GLPK, CBC), Deep Learning (PyTorch, Transformers), OpenCV, Scikit-learn

Positions of Responsibility & Extracurricular Activities

  • Deep Learning TA (IE 643): Mentored students to help with conceptual, technical, and implementation-related doubts.
  • Sponsorship Coordinator (IEOR): Facilitated industry-student collaborations for technical projects and corporate ties.
  • Interview Coordinator (IIT Bombay): Helped place 2000+ students, organized and managed assessment tests for 100+ firms.
  • Teaching Assistant (Communication Skills): Mentored 100+ students in communication, presentations, and technical writing; guided research papers, PPTs, and Excel; conducted workshops, achieving a 20% skill improvement.
  • Kaggle Rank: Achieved 6th rank in the IE 506 Machine Learning course competition, with an F1 score of 0.869 using XGBoost.
  • Hobbies: Passionate about gym workouts, watching Marvel movies, exploring the latest tech gadgets, and discovering new cuisines.

Interests

  • Teaching the next generation of data scientists
  • Exploring the intersection of AI and real-world problems
  • Writing blogs on Data Science and AI/ML
  • Staying fit with gym workouts
  • Analyzing Marvel movie plots