Machine Learning Project - Long Term Short Term Model [In Progress]
Machine Learning Project - Simple Pytorch Learning Model April 2025 Link

• Built a fully functional machine learning model from scratch using PyTorch’s nn.Linear layer, demonstrating an understanding of model architecture, tensors, and data flow in neural networks.
• Designed and implemented a custom training loop including forward propagation, loss calculation with nn.MSELoss, backpropagation, and weight updates with torch.optim.SGD.
• Applied Mean Squared Error (MSE) loss function to measure model performance and optimize model parameters over multiple epochs to minimize prediction errors.
• Demonstrated understanding of optimizers and gradient descent by manually updating model parameters based on calculated gradients to improve model accuracy over time.
• Reduced model loss by over 80% over the training period, showing successful learning and convergence through iterative optimization.
• Explained complex machine learning concepts (loss functions, optimizers, training loops) through a storytelling analogy ("Bob the Model and Coach Smith") to make foundational AI/ML principles more accessible to beginners.
• Documented the project with a detailed README explaining the project goals, learning outcomes, model architecture, and future improvement plans, ensuring clarity for technical and non-technical audiences.

Machine Learning Project - Simple Learning Regression Model April 2025 Link

• Built a supervised machine learning model to forecast daily maximum temperatures using historical weather data from Kaggle.
• Applied time series feature engineering techniques, including lag features, rolling averages, and seasonality transformations.
• Trained and evaluated a Stochastic Gradient Descent(SGD) forecasting model, achieving an RMSE (Root Mean Squared Error) as low as 1.61 on test data.
• Conducted rigorous model validation using Time Series Cross-Validation, reporting an average RMSE of 3.13 across multiple folds.
• Engineered seasonal features (day-of-year sine/cosine cycles) to improve model accuracy and capture periodic patterns.
• Developed and visualized forecast vs actual temperature trends using matplotlib, ensuring clear model performance interpretation.
• Gained hands-on experience with key machine learning libraries such as scikit-learn, pandas, numpy, and matplotlib.
• Demonstrated strong understanding of time-series forecasting challenges, including data leakage prevention (no shuffling) and temporal dependencies.

Machine Learning Project - Simple Tensor Model April 2025 Link

• Designed and implemented a simple TensorFlow model to perform supervised learning tasks, demonstrating strong understanding of tensor operations and data flow within neural networks.
• Applied core mathematical principles to understand and manipulate tensors, activation functions, and gradient-based optimization techniques during model training.
• Built and trained a basic model architecture from scratch, including defining input/output layers, weights, and bias initialization without reliance on pre-built templates.
• Optimized model performance by analyzing training loss behavior and understanding gradient descent convergence through hands-on experimentation.
• Demonstrated deep understanding of foundational machine learning concepts by translating mathematical formulas into functional TensorFlow code.
• Improved technical problem-solving skills by debugging tensor operations, dimensionality errors, and learning rate tuning challenges.
• Documented project with clear explanations of the underlying mathematics, model structure, and training results for easy reproducibility and future development.
• Strengthened Python programming skills with a focus on TensorFlow library usage, object-oriented design, and data manipulation.

Data Analyst Project - United States Crude Oil Import & Export Report Sep 2023 - Oct 2023 Link

• Demonstrated reporting on research United States' top crude oil imports and exports for fiscal year 2022 self-learning project, for reporting on U.S Top Exports of Oil and Gas, turning raw data to identifying meaningful information, for strategic planning
• Create professional, visually appealing market trends, including charts, graphs, and infographics, from multiple source systems into data visualization report (Power BI).
• Detail oriented with excellent organizational, reporting solution, analytical, and problem-solving skills using DAX calculations and data modeling.
• Assessed the report to identify significant trends or changes over time, contributing to a comprehensive understanding of market.

Data Analyst Project - U.S Sales of EVs & PHEV Link

Excel Project - Self Project (Sept 2022 - Oct 2022)

• Crafted drill down reports on weather data using Pivot tables, leveraging sorting, grouping, filtering, and slicing techniques to distill complex data.
  
• Designed accompanying charts and diagrams to visually communicate trends and patterns in data.
  
• Enhanced overall understanding and informing strategic decision-making.
  
• Showcasing proficiency in data analysis, visualization, and attention to detail
  

Senior Project: Sentio AI: Stocks Sentiment Analysis Project The University of Houston – Victoria (Augest 2022 - December 2022)

• Assisted the Lead Team member in developing an iOS app utilizing Apple's ML Core and Twitter Elevated+ API.
  
• Provided guidance to peers on API implementation and functionality.
  
• Gained experience in programming with Apple's Swift language.
  
• Developed proficiency in Git version control system.
  
• Collaborated with team members on assigned tasks and shared responsibilities.
  

Database System: Coffee House Database The University of Houston – Victoria (Augest 2020 - Dec 2020)

• Demonstrated effective leadership skills as the team lead and resolved team issues, guiding a group of 3 members towards successful project completion.
  
• Leveraged expertise in Android Studios, PHP, MySQL, and Java to develop and deploy an end-to-end 3-tier database system, ensuring seamless functionality throughout the entire stack.
  
• Designed a robust ER-Model for the database system, showcasing meticulous attention to detail and a strong understanding of database design principles.