// Research Portfolio
Building at the intersection of Deep Learning, Computer Vision, and NLP.
Investigates multiple state-of-the-art transfer learning architectures for the automated detection and classification of diseases in date palm leaves across multiple infection categories, enabling precision agriculture through deep visual recognition.
Benchmarks leading transfer learning models for fine-grained visual classification of pistachio species, contributing to automated agricultural quality control and botanical species recognition using convolutional feature extraction.
Applies Fully Convolutional Networks to the nuanced task of identifying ragas in Indian classical music, bridging deep learning with ethnomusicology for automated music analysis and cultural heritage preservation.
Proposes an ensemble-based classification framework for Coronary Artery Disease detection, combining attribute selection and elimination strategies to improve diagnostic accuracy in clinical decision support systems.
Examines the integration of deep learning architectures for financial time-series forecasting, evaluating model capacity to capture complex market dynamics for improved stock price prediction accuracy.
Provides a comprehensive comparative analysis of classical machine learning and modern deep learning models for stock price prediction, evaluating predictive performance and model robustness across diverse market conditions.
Develops an NLP-driven pipeline using Naïve Bayes classification to automate the identification of cyberbullying in online text, addressing online safety through intelligent content moderation and harm detection systems.
Examines machine learning approaches to forecast Uber trip durations while incorporating weather and traffic flow variables, advancing urban mobility intelligence and data-driven transportation planning.
I'm Shubbh Rajesh Mewada, a researcher and engineer working at the intersection of Machine Learning, Deep Learning, Computer Vision, and NLP.
My published work spans medical AI, agricultural image classification, music signal processing, financial forecasting, and online safety, with 105 citations across 8 conference papers, all published in 2023.
Currently building systems that push the boundaries of what intelligent models can do in the real world.