Saurav Singh Chandel

About Me

I'm Saurav Singh Chandel, a driven fourth-year student at Memorial University of Newfoundland pursuing a joint honors degree in Computer Science and Applied Mathematics. Problem-solving is one of my key strengths, and I'm passionate about leveraging my analytical skills and technical expertise to create innovative solutions. I'm particularly drawn to the exciting possibilities within AI/ML and backend development, and I'm eager to explore these areas further.

Currently, I'm conducting my honors research, investigating whether a surface is a marginally outer trapped surface using spectral analysis in the context of Black Holes. Additionally, I've honed my skills through projects like building a YouTube clone, showcasing my proficiency in cloud-native architectures, serverless deployment, and modern frontend development with Next.js and TypeScript. I thrive on challenges and am dedicated to continuous learning and growth in the ever-evolving field of technology.

Linked In: www.linkedin.com/in/saurav-singh-chandel-92937221b

GitHub: https://github.com/SauravSinghChandel/

E-mail: sschandel@gmail.com

Projects

The Message Board is a Python-based web application designed to foster online discussions and community interaction. It features an interface for creating posts, showcasing my ability to develop interactive web applications that facilitate communication and collaboration.

I've developed a YouTube clone that leverages a cloud-native architecture using Google Cloud Run for serverless deployment, Pub/Sub for efficient message queuing, and Storage buckets for scalable media handling. The backend logic is powered by Firebase Functions, while the frontend provides a modern, type-safe user experience built with Next.js and TypeScript.

StockUP is a full-stack JavaScript web application that simulates a fast-paced stock trading environment. Users race against the clock to reach a target portfolio value by strategically buying and selling stocks using real-time data from the Polygon API. This project demonstrates my ability to build dynamic, data-driven web applications with engaging user experiences.

Research Papers

This study explores Reinforcement Learning (RL), using Stable Baselines to implement agents like DQN and PPO for decision-making in diverse states. We train agents for the Atari Space Invaders environment, experimenting with training steps and visualizing performance metrics. This research contributes to understanding how machines learn and make intelligent decisions.
Link to the video: Demonstration Video

My research investigates the optimization of neural networks, visualizing their behavior and exploring the impact of architecture and optimization strategies on performance. This work aims to provide valuable insights for developing more effective neural network applications.

This research navigates the dynamic field of machine learning optimization, focusing on the design and empirical evaluation of Python-based algorithms for both single and two-parameter loss functions. The findings provide crucial insights and practical guidance, empowering the machine learning community to make informed decisions for optimizing model performance in the ever-evolving landscape of artificial intelligence.

Note: These research papers were written as a part of a course