Pinned Repositories
Homesite_quote_conversion_prediction
This repository contains a data science case study on the 'Homesite quote conversion' problem from Kaggle. The objective of the project is to develop a predictive model that accurately identifies potential customers likely to convert on a Homesite insurance quote.
LLM-From-Scratch-For-ChatBots-GPT2
This project implements an LLM from scratch, utilizing the transformer mechanism similar to that of GPT-2. The model is trained on a customer service dataset.
Backpropagation_Custom_Implementation
This repository contains a Python implementation of a neural network optimization process using different optimizers: Vanilla, Momentum, and Adam. The neural network's forward propagation, backward propagation, gradient checking, and optimization process are included in the code.
Custom_SGD_Logistic_Regression
This repository contains a custom implementation of the Stochastic Gradient Descent (SGD) classifier for binary classification. The classifier is implemented with Logarithmic Loss (Logloss) and L2 regularization, all without relying on the scikit-learn library.
Self_Driving_Car_Village_Roads
This repository contains code for a self-driving car project trained on village road scenarios. The model predicts steering angles based on input images, and the simulation includes a visual representation of the predicted steering angle on a virtual steering wheel.
YOLOv8_Custom_Dataset_Pothole_Detection
This repository implements a custom dataset for pothole detection using YOLOv8. The project focuses on training and fine-tuning YOLOv8 on a specialized dataset tailored for pothole identification. The code includes training scripts, pre-processing tools, and evaluation metrics for quick development and deployment.
Implementing_Callbacks
This GitHub repository contains a project that focuses on experimenting with various callback functions in TensorFlow for enhancing model training and performance in a neural network-based binary classification task. The project involves constructing multiple neural network architectures, optimizing hyperparameters, and assessing model performance.
Sentiment-Analysis-System
A web-based NLP tool that predicts sentiment (Positive, Negative, Neutral) from text, with visualizations by demographics and location. Integrating Google Forms for input and VADER for analysis, it uses Pandas and Plotly for data processing and visualization. Built with Streamlit and Python, accessible via LAN.
Vehicle_Insurance_Claims_Prediction
This repository contains code for predictive models in vehicle insurance claims analysis. It includes data preprocessing, feature engineering, and machine learning techniques. The repository aims to provide insights and solutions for improving insurance claim predictions.
Amazon-Review-Sentiment-Analysis-System
The Amazon Review Sentiment Analysis System is a deep learning-based model that predicts sentiments from review data and integrates with the Google Sheets API for data management. The application presents results using Streamlit and visualizations created with Plotly.
AnoopCA's Repositories
AnoopCA/Face-Mask-Detection-System
A real-time application that detects face mask usage via video streams using TensorFlow for model training and OpenCV for video processing. The Streamlit-based interface supports images, videos, webcam, and IP camera feeds, ensuring public safety and compliance with health protocols.
AnoopCA/Customer-Management-System
This is a streamlined application for managing customer data, transactions, and interactions. It includes features like tracking purchases, managing feedback, and handling inquiries, while enabling targeted marketing campaigns. Built with Streamlit, Python, and MySQL, it is accessible via web browsers.
AnoopCA/Drowsiness-Detection-System
This project implements a Drowsiness Detection System to monitor a person's alertness while driving or working by detecting prolonged eye closure using a webcam or video feed. It utilizes a pre-trained TensorFlow model and OpenCV's Haar Cascade for real-time eye detection and classification.
AnoopCA/Self_Driving_Car_Village_Roads
This repository contains code for a self-driving car project trained on village road scenarios. The model predicts steering angles based on input images, and the simulation includes a visual representation of the predicted steering angle on a virtual steering wheel.
AnoopCA/YOLOv3_Custom_Implementation
This GitHub repository hosts an implementation of YOLOv3 (You Only Look Once version 3) from scratch, designed to provide a comprehensive understanding of this state-of-the-art object detection algorithm. YOLOv3 is renowned for its real-time object detection capabilities, making it a popular choice for a wide range of computer vision applications.
AnoopCA/Hangman_Letter_Guessing_Game
This Python repository hosts a Hangman Letter Guessing Game, where users can experience the classic word-guessing game and witness the AI's letter guessing capabilities. The user has the flexibility to choose the number of game rounds and observe the outcomes.
AnoopCA/YOLOv8_Custom_Dataset_Pothole_Detection
This repository implements a custom dataset for pothole detection using YOLOv8. The project focuses on training and fine-tuning YOLOv8 on a specialized dataset tailored for pothole identification. The code includes training scripts, pre-processing tools, and evaluation metrics for quick development and deployment.
AnoopCA/Custom_VGG16_Document_Image_Classification
This is a data science project that leverages a custom VGG16 deep learning model to accurately classify document images into predefined categories, with applications in document management and content organization.
AnoopCA/Sentiment-Analysis-System
A web-based NLP tool that predicts sentiment (Positive, Negative, Neutral) from text, with visualizations by demographics and location. Integrating Google Forms for input and VADER for analysis, it uses Pandas and Plotly for data processing and visualization. Built with Streamlit and Python, accessible via LAN.
AnoopCA/Amazon-Review-Sentiment-Analysis-System
The Amazon Review Sentiment Analysis System is a deep learning-based model that predicts sentiments from review data and integrates with the Google Sheets API for data management. The application presents results using Streamlit and visualizations created with Plotly.
AnoopCA/Employee-Management-System
An Employee Management System (EMS) with a user-friendly Streamlit frontend and MySQL backend. Features include employee information management, attendance tracking, leave management, payroll processing, and project assignments.
AnoopCA/LLM-From-Scratch-For-ChatBots-GPT2
This project implements an LLM from scratch, utilizing the transformer mechanism similar to that of GPT-2. The model is trained on a customer service dataset.
AnoopCA/Vehicle_Insurance_Claims_Prediction
This repository contains code for predictive models in vehicle insurance claims analysis. It includes data preprocessing, feature engineering, and machine learning techniques. The repository aims to provide insights and solutions for improving insurance claim predictions.
AnoopCA/TFIDF_Custom_Implementation
This repository contains implementations of the Term Frequency-Inverse Document Frequency (TF-IDF) feature extraction technique using both SkLearn library and a custom implementation.
AnoopCA/Social_network_Graph_Link_Prediction
This GitHub repository contains code and analysis for the Social Network Graph Link Prediction project, specifically focused on the Facebook Challenge. The goal of this project is to predict whether a link between two nodes (users) will be established in a social network graph.
AnoopCA/Random_Search_CV_Custom_Implementation
This repository contains a custom implementation of the Randomized Search Cross-Validation (RandomSearchCV) algorithm for hyperparameter tuning, specifically focusing on the hyperparameter K for the K-Nearest Neighbors (KNN) classifier.
AnoopCA/Performance_Metrics
This repository contains custom implementations of various performance metrics for evaluating classification and regression models. The metrics are calculated and demonstrated on different datasets.
AnoopCA/Naive_Bayes_on_DonorsChoose_dataset
This repository contains code for predicting whether a project on the DonorsChoose platform will be successfully funded or not using a Naive Bayes classifier. The DonorsChoose dataset provides information about educational projects and their funding outcomes.
AnoopCA/Implementing_Callbacks
This GitHub repository contains a project that focuses on experimenting with various callback functions in TensorFlow for enhancing model training and performance in a neural network-based binary classification task. The project involves constructing multiple neural network architectures, optimizing hyperparameters, and assessing model performance.
AnoopCA/Document_Classification_with_CNN
This GitHub repository showcases a document classification project using Convolutional Neural Networks (CNN). The aim of the project is to classify documents into different classes using the CNN model.
AnoopCA/Custom_SGD_Logistic_Regression
This repository contains a custom implementation of the Stochastic Gradient Descent (SGD) classifier for binary classification. The classifier is implemented with Logarithmic Loss (Logloss) and L2 regularization, all without relying on the scikit-learn library.
AnoopCA/Clustering_on_Movie_Actor_Network_Dataset
This GitHub repository contains code to group similar movies and actors using clustering techniques. The project utilizes network analysis, node embedding, and clustering algorithms to identify and visualize clusters of similar movies and actors.
AnoopCA/Backpropagation_Custom_Implementation
This repository contains a Python implementation of a neural network optimization process using different optimizers: Vanilla, Momentum, and Adam. The neural network's forward propagation, backward propagation, gradient checking, and optimization process are included in the code.
AnoopCA/Homesite_quote_conversion_prediction
This repository contains a data science case study on the 'Homesite quote conversion' problem from Kaggle. The objective of the project is to develop a predictive model that accurately identifies potential customers likely to convert on a Homesite insurance quote.