MediClassify: A High-Accuracy Multimodal Classification System for Medical Paper Abstracts

• Engineered and trained a multimodal classifier combining NLP and structured data approaches using TensorFlow and Scikit-Learn, achieving 96% accuracy in classifying medical paper abstracts
• Employed BERT embeddings, ensemble methods to improve robustness and accuracy for complex medical text analysis

Model Architecture