Data cleaning is done using regex expressions. Logistic regression with tfidf is applied that got 68 percentage accuracy on test data. Neural network is trained using Keras sequential model using pre trained embedding that got 83 percentage est accuracy.

Project is the Natural Language Processing(NLP) related problem. NLTK library is used to clean the data.
The solution approach achieved a good classification accuracy of 0.92 using transfer learning. The ULMFit
model is trained on the dataset using fast-ai library.

Developed LLM chatbot by fine-tuning from financial news data, which answers user query using RAG and update the Vector DB realtime from news API.