🔭 I’m currently working on Utilizing variants of Graph neural network for intrusion detection models for advanced persistent threats.

🌱 I’m currently learning Graph neural networks, RAG based LLM's

📫 How to reach me [email protected]

📄 Resume Drive Link

• Developed a high-accuracy binary classification model using TensorFlow and Keras for differentiating cat and dog images.
• Enhanced model performance by 20% using data augmentation techniques like rotation and flipping.
• Fine-tuned hyperparameters (dropout rates, filter sizes) to improve training efficiency, reducing time by 15%.

• Built a predictive model for rental prices using feature engineering techniques like frequency encoding and missing value handling.
• Explored multiple algorithms (Linear Regression, Random Forest, Gradient Boosting) to optimize prediction accuracy.
• Achieved a 10% improvement in price prediction accuracy through hyperparameter tuning and model evaluation.

• Curated a balanced dataset from ImageNet-1k and built a CNN-based classifier for food vs. non-food detection.
• Achieved a 93.67% training accuracy and 87.28% validation accuracy, with enhanced model stability through Batch Normalization and Dropout.
• Used Weights & Biases to monitor training progress and optimize model performance in real-time.

• Developed a hybrid Graph Convolutional Network (GCN) and LSTM model for classifying paranoid schizophrenia from EEG data.
• Engineered advanced graph-based features and optimized temporal patterns using LSTM layers, achieving 97% accuracy.
• Implemented robust cross-validation testing to ensure consistent model performance across diverse datasets.