Research

Pneumonia Disease Prediction

• Paper Titled with “Pneumonia Disease Prediction using VGG19 Architecture”. 

• Accepted and Presented at Hybrid Intelligent Systems(2023) - A Springer International Conference.

• Novelty  is to use the VGG19 Transfer Learning Model along side Callback  API's like Early Stopping, , ReduceLROnPlateau to reduce the overfitting and alter the Learning Rate to increase the efficiency of the model respectively.

Financial Complaint  Summarization

• An Extensive Research has been done to summarize the text more optimistically, using the Large Language Model's (LLM's) and Reinforcement Learning Approaches.

• Our Idea is build a Hybrid Summarization model using Hugging Face  that can be further optimized using Reinforced Q-Learning Approach.

• The Hybrid Model includes Extractive and Abstractive Summarization models, where BERT and T5 Summarization models used respectively.

• The Quality of Summarization Models are evaluated using ROUGE-Score, and the models are performed exceptionally well with a scores of 90.60(Extractive) & 80.40 (Abstractive).