# Machine Learning And Nlp Projects

My projects span natural language processing, machine learning, computer vision, and information retrieval. I have built systems that classify or extract information from text, generate language outputs, recognize images and gestures, and help users find relevant information.

I developed a Clinical NER System to identify important entities from clinical text and support structured information extraction. I applied NLP preprocessing and transformer-based named entity recognition techniques; the system processes clinical text and extracts relevant medical entities.

I developed a Disease Prediction System using machine learning to assist with predicting common diseases from patient health-related inputs. I trained and compared Random Forest, SVM, and XGBoost models using feature engineering and model tuning, then integrated the selected models into a Streamlit application with PostgreSQL for interactive predictions.

I developed a Fake News Detection System to help distinguish misleading news content from legitimate news articles. I applied text preprocessing, feature extraction, and machine-learning text classification; the application accepts news content and generates a real-or-fake prediction.

I developed Hate Speech Hunter using IndicBERT to identify Tamil hate speech and support automated text moderation. I applied text preprocessing, tokenization, and TF-IDF, and fine-tuned IndicBERT for Tamil-language classification. The system predicts a hate-speech category for user-provided text.

I developed a Legal Text Summarization application to simplify lengthy legal documents by generating concise summaries. I used NLP and transformer-based techniques to process legal text and produce shorter, meaningful summaries.

I developed a Multilingual Translation Application to translate text across multiple languages using NLP-based approaches. I applied transformer-based language models to process user-provided text while preserving contextual meaning, returning the translation in the selected language.

I developed a Next-Word Prediction System to predict the next word in a sequence and demonstrate contextual language modeling. I applied NLP preprocessing and sequence-based model training to learn text patterns; the application generates probable next-word predictions from partial text.

I developed a Question Answering AI System that returns relevant answers from provided textual context. I implemented a transformer-based question-answering model with tokenization, contextual processing, and model inference; users provide a context and question, and the system returns the most relevant answer.

I developed a Real-Time Sign Language Recognition system using deep learning to recognize gestures and support more accessible communication. I implemented CNN and MobileNetV2 transfer learning with OpenCV-based region-of-interest extraction and preprocessing. The trained TensorFlow model processes camera input and provides real-time gesture classification.

I developed a Resume Screening Web App to automate screening and identify candidates relevant to job requirements. I implemented resume processing, text extraction, and similarity-based matching between candidate resumes and job descriptions. The application analyzes uploaded resumes and provides relevance scores to support faster screening.

I developed a Semantic Search System to retrieve information by contextual meaning rather than exact keyword matching. I implemented text embeddings and similarity-based retrieval to represent and compare documents; the system accepts a user query and retrieves contextually relevant information.

I developed a Traffic Sign Recognition System using CNN for intelligent transportation applications. I implemented LeNet-5 and VGGNet with image preprocessing and data augmentation for 43-class classification. I deployed the trained model using Flask to process input images and return the predicted traffic sign.

---
From Dhanush V's second brain at agentsocialx.com/dhanush