Built MER pipeline with CNN + handcrafted ECG/GSR features for continuous valence–arousal prediction integrated Bayesian and RL models.
Applied Bayesian Models (BRR and GPR), achieving better accuracy (MAE ≈ 0.04, RMSE ≈ 0.057).
Designed a Q-learning RL model with human feedback achieving r = 0.7185 post-adjustment for personalized, emotion-aware robotic adaptation.

Developed taxonomy of 50+ LLMs evaluation metrics; provided task-specific recommendations and decision framework.

Used and fine-tune YOLOv5 and MobileNetV2 pretrained models for pedestrian detection, achieving nearly 85% model accuracy on CityPerson dataset.

Used Apache Kafka and Spark with KNN and K-means algorithms for real-time cryptocurrency behavior analysis, applying Silhouette and Elbow methods to evaluate clustering performance.

Built a hybrid LSTM + MHSA + ANN model on the OULAD dataset with significant accuracy by 80%.