Karan_Portfolio

My GitHub Portfolio

Welcome to my GitHub portfolio! Here, you’ll find a collection of my projects, showcasing skills in machine learning and analytics. Each project highlights a unique challenge I have tackled, demonstrating my problem-solving and technical abilities.

Project Index

Project Overviews

Canadian Innovation Strategy - An End-to-End Analysis

image

Description: Innovation is the cornerstone of economic growth and prosperity, acting as a catalyst for transforming economies and industries. Despite Canada’s commendable position within the top 20 innovative countries globally, there remains significant potential for enhancement. Our project undertakes a comprehensive analysis to chart a path for advancing Canada’s innovation landscape. By harnessing a variety of data analysis tools and advanced data science methodologies, we dissect patterns of success from leading nations to formulate a robust innovation strategy. This end-to-end project not only identifies key areas for improvement but also provides actionable insights and recommendations, leveraging global best practices to foster a more innovative Canada.

Technologies: Python, Matplotlib, ScikitLearn, ChatGPT API, Pandas

View Project

Natural Disaster Alert System

5

Description: In today’s fast-paced world, staying informed about natural disasters can be a challenge, especially with our busy schedules and the global spread of loved ones. Our application bridges this gap by providing timely notifications about impending natural disasters in your designated areas of interest. This system leverages advanced monitoring technologies to alert users about potential threats, ensuring you and your loved ones stay informed and prepared, no matter where you are in the world.

Technologies: AWS (Lambda, DynamoDB, API Gateway), JavaScript, HTML, Python

Automated Notification Email:

image

Backend Flow:

Screenshot 2024-03-12 173010

Code on AWS

YouTube Market Insight Summarizer

image

Description: In the dynamic world of stock investing, staying informed is key, yet the abundance of lengthy, often lengthy YouTube videos on market trends and stock recommendations poses a challenge for the casual investor. With time as a precious commodity, sifting through these extensive videos for valuable insights becomes a daunting task. The solution bridges this gap by offering a comprehensive system designed to summarize these videos, extracting and presenting the most pertinent information in a concise, user-friendly format. Utilizing the YouTube API for transcript extraction and the capabilities of the ChatGPT API for content summarization, this project efficiently condenses the core insights from extensive market analysis videos. This approach allows investors to stay updated on market trends and make well-informed decisions, eliminating the need to sift through hours of footage. This solution harnesses these technologies to streamline the process of gathering crucial investment information, making it accessible and time-efficient for casual investors

Technologies: Python, Chat-GPT API, MatplotLib, Dash, HTML

Step 1: Input Youtube Video ID

Sample URL: https://www.youtube.com/watch?v=BWrkByUnfEk&ab_channel=FinTek

STEP 2: Click the Search button

image

Step 3: Select Stock of Interest and View Concise Report image

View Project

Next Steps: Update the front end of the solution. Create UI to take input URL and output concise stock information accordingly. Add more variety in terms of information presented.

Image Resolution Enhancer - In Progress

image

Description: Blurry images can significantly diminish the impact and usability of visual content, whether in professional presentations or personal memories. Addressing this common yet challenging issue, our project introduces a powerful image resolution enhancer powered by deep neural networks. Designed with versatility in mind, this application has been trained on a wide-ranging dataset, enabling it to effectively restore and enhance the quality of images across various contexts. From sharpening personal photos to refining professional graphics, our tool offers a seamless solution to improve image clarity and detail, ensuring your visuals are always of the highest quality.

Technologies: Python, PyTorch, Scikit-Learn, Numpy, Matplotlib

GitHub Link - Available Soon

NYC Real Estate Predictor

image

Description: This project is a NYC real estate price predictor based on 2024 historical prices. The dataset obtained was carefully analyzed and cleaned for modeling. The dataset was subject to a variety of models such as linear regression and XGBoost models. Upon hyperparameter tuning the model was deployed onto a Flask server to enable RESTful API collaboration, providing real-time house price predictions to the front-end website.

Technologies: Python, Scikit-Learn, XGBoost, Flask, JavaScript, HTML

GitHub Link - Available Soon

About Me

Master of Engineering student at the University of Toronto with a strong academic foundation in data science, statistical analysis, and machine learning. A quick learner and detail-oriented problem solver with a passion for utilizing data to drive insights and make informed decisions. Looking for Full-Time Data positions starting May 2024.

Contact Me:

Thank you for visiting my portfolio!