Diwash Thapa
Undergraduate Research Assistant specializing in Machine Learning and AI. Building intelligent systems for particle detection and turbulence analysis.
About Me
Passionate about leveraging AI and Machine Learning to solve complex problems
Education
Expected Dec 2027
Texas Tech University
Bachelor of Science in Computer Science
Research Focus
Current Research
Working on machine learning models for particle detector data analysis and turbulence prediction using deep neural networks.
Interests
Neural Networks, Deep Learning, Data Science, Predictive Analytics, and AI applications in Physics and Engineering.
Conference
Presented at URC 2025: "Using Deep Neural Network for Predicting Small-Scale Dynamics of Turbulent Flow"
Skills & Expertise
A comprehensive toolkit for building intelligent systems
Technical Skills
Data Science
Tools
Work Experience
Building expertise through research and real-world applications
Research Assistant
Computational Intelligence, Control and Information Lab
- •Researching advanced neural network optimization techniques to improve learning efficiency and model robustness.
- •Developing novel activation functions to address neuron death and the vanishing gradient problem.
- •Collaborating with graduate researchers and faculty to advance deep learning architectures.
Data Intern
FRPS Investment
- •Analyzed sales and inventory data to identify trends and develop reports and dashboards for data-driven decision-making.
- •Collaborated with marketing and sales teams to integrate data-driven strategies.
- •Enhanced the performance of e-commerce and inventory management tools alongside senior developers.
Undergraduate Research Assistant
Turbulence and Big-Data Lab
- •Assisted in the development and application of machine learning models for analyzing complex turbulence datasets.
- •Collaborated with professors to optimize algorithms, improving the accuracy of data-driven insights.
- •Conducted data preprocessing, feature engineering, and model evaluation to support predictive analytics research.
Selected Projects
Hands-on experience building ML models from scratch
Handwritten Digit Recognition
2025Built a multi-layer neural network from scratch using NumPy to classify MNIST handwritten digits, achieving 96% test accuracy without high-level ML libraries.
Key Highlights:
- ▸Implemented custom one-hot encoding function for multi-class classification
- ▸Designed manual gradient descent and backpropagation logic
- ▸Validated with gradient checking to ensure correctness
Power Consumption Analysis
2024Collaborated on a data analysis project examining power consumption across three zones in Tetuan, Morocco, considering weather, landscape, and occupancy factors.
Key Highlights:
- ▸Analyzed 2017 power consumption data to identify trends
- ▸Developed three best-fitted predictive models
- ▸Enhanced understanding of energy needs across urban zones
F1 Real-Time Race Strategy Analysis
2025Designed and deployed a real-time machine learning system using Bayesian Neural Networks to analyze live Formula 1 race data and optimize race strategies with over 90% prediction accuracy.
Key Highlights:
- ▸Trained Bayesian Neural Networks achieving 90%+ accuracy on live F1 race data
- ▸Built an end-to-end data pipeline integrating Python ML workflows with a Next.js dashboard
- ▸Performed advanced data preprocessing including missing value handling, normalization, and feature engineering
Achievements & Awards
Recognition for academic excellence and research contributions
Conference Presentation
Undergraduate Research Conference 2025
Presented poster titled "Using Deep Neural Network for Predicting Small-Scale Dynamics of Turbulent Flow"
April 2025AI Fellowship
Break Through Tech
Selected for the Breakthrough Tech AI Program, an industry-driven AI Training initiative
May 2025 - April 2026CISER Scholars Program
Texas Tech University
Recognized for academic excellence and research potential
August 2025TrUE Scholars Program
Texas Tech University
Selected for undergraduate research excellence program
July 2024Get In Touch
I'm always open to discussing research opportunities, collaborations, or just connecting!