mumin

Recent Projects

Leveraging GPT-4 for Enhanced SQL Query Generation Using Natural Language Prompts

We tried to understand which additional contexts yield better one-shot prompt to SQL results on GPT-4.

Learn more.

Irrigation Ablation Study

Our team focused on agricultural irrigation detection and fraud detection due to the significant impact irrigation has on global water resources. We utilized the BigEarthNet dataset to perform an ablation study, aiming to identify which features most influence machine learning models. This study helps future researchers quickly pinpoint impactful features for their models.

Learn more.

NIH Chest X-Ray Classification

This project focused on classifying the NIH chest x-ray dataset using various deep neural networks. We optimized the model through hyperparameter tuning and different architectures, aiming to replicate or improve results from the "ChestX-ray8" paper. The dataset included over 100,000 anonymized images from more than 30,000 patients.

Learn More