Parv Thakkar's Projects
This project successfully developed a deep learning model using bidirectional LSTM, achieving an impressive 83% test accuracy and 96% train accuracy in classifying Stack Overflow questions into high-quality, low-quality requiring edits, and low-quality closed categories.
An advanced AI Face Recognition system using Convolutional Neural Networks, showcasing high accuracy and innovative implementation for diverse applications.
Developed a comprehensive Tableau dashboard to analyze Airbnb rental prices, effectively visualizing average prices, price variations by number of bedrooms, and pricing trends across different zip codes.
Comprehensive database project for analyzing bird strike incidents on aircraft, involving data loading, SQL querying, and analytics using MySQL/MariaDB and R.
Machine learning-driven project for predicting Black Friday sales, providing insights into consumer behavior and enhancing business strategies.
A Blackjack Card game created in python using tkinter.
An analytical exploration of Boston's crime data, using EDA and time-series modeling to predict crime patterns and aid in better decision-making for public safety.
Advanced image classification project on the CIFAR-10 dataset, leveraging deep learning techniques to distinguish between various image categories.
An insightful Power BI dashboard providing detailed analysis of cricket players' performances, designed for cricket analysts and enthusiasts.
A user-friendly GUI system for efficient hotel reservation management, built to enhance the booking experience for hotels and customers alike.
A professional portfolio website showcasing personal details, career achievements, and educational background.
Developed three predictive models using the VitalDB dataset to forecast patient diagnoses, ICU length of stay, and mortality rates. Improving the accuracy of the patient diagnosis model by 10%, and achieving high accuracy rates with models for ICU stay (83%) and mortality (96%).
A detailed project demonstrating the extraction, transformation, and loading of XML data into relational and analytical databases, using SQLite and MySQL.
An innovative AI-powered system to monitor and enforce social distancing in public and office spaces, utilizing advanced computer vision techniques.
In this project I have analyzed various stocks of different companies and commodities. Using matplotlib, scikit-learn, numpy, etc.
I executed an End-To-End Data Engineering Project on Real-Time Stock Market Data using Kafka.
Analyzing and predicting student performance through data-driven insights and machine learning models, aimed at enhancing educational outcomes and strategies.
Developed a data pipeline in Azure to process Tokyo Olympics data, involving data ingestion, transformation with Databricks and PySpark, and analysis using Azure Synapse Analytics and SQL.
Developed a focused web crawler for Apple technology content, collecting 4000 documents into an Elastic Search cluster with prioritization based on keywords and in-link counts, using Beautiful Soup for data processing.
Efficiently collecting data from the Twitter API and storing it in a GCS bucket. With a custom Python script ensuring data reliability, the pipeline streamlined the process, enabling accessible and dependable Twitter insights.