• Created a Credit card fraud detection model using PCA applied open-sourced dataset by ULB.
• Performed Data Visualization to check outliers and remove them using Tukey’s IQR method .
• Resampling our imbalanced dataset using random oversampling and used Random Forest algorithm to train ourmodel.
• Used SMOTE, Tomek links to resample the dataset and trained model with Random Forest.
• Found out the best hyperparamters using GridSearchCV for each of the resampling techniques.
• Used recall score as metrics for evaluating our model performance for each resampling technique.

• Developed an application that allows us to chat with multiple PDF documents at the same time.
• Extracted Contents of multiple PDF documents and divided them into smaller text chunks for easy processing.
• Used Large Language Model to generate embeddings of the text chunks and stored them in vector database.
• Used Faiss vector database for similarity search of the vectors which are more semantically similar .
• Passed the text to language model to generate the response for the user.
• Future scopes of the project will be to extend the input to image and audio files and process user query upon it.

Lorem ipsum, dolor sit amet consectetur adipisicing elit. Molestiae necessitatibus error eos suscipit culpa, impedit cum corporis dignissimos porro quae deleniti. Mollitia blanditiis, doloremque eius voluptas nam labore eligendi maiores in sed pariatur odit provident qui ad est similique iste. Lorem ipsum dolor sit amet consectetur, adipisicing elit. Sint assumenda voluptatibus eligendi excepturi blanditiis molestias minima labore officiis fugit aliquam. Lorem ipsum dolor sit amet, consectetur adipisicing elit. Optio dolores quos atque reprehenderit qui esse, autem ab sequi nihil facilis?