● Researched and created a custom Marathi Word Sense Disambiguation Application using PyQt5, Natural Language Processing, and Deep Learning Techniques.
● Researched and created a medical document classification and data extraction system / Flask API that uses a hybrid approach for document classification (with 98.8% F1-Score), and Bi-LSTM for entity extraction (with 98.6% F1-Score).
● Created a Waste Classification system using transfer learning techniques and Machine Learning.
● Performed artificial anomaly insertion and Anomaly Detection on 21 days of Satellite Data, collected at each second of the day, using LSTM.
● Skills: Python, Tensorflow, Keras, Pytorch, PyQT5, Flask, CNN, LSTM, Git, Named Entity Recognition