Portfolio

Deep Sleep AI

Engineered a fully-automated lightweight-performance sleep stage classification system for real-time electroencephalography signal processing
Developed a custom Python library to extract statistical features from EEG signals using high-performance computing (CUDA, CuPy)
Implemented transformers to track sleep stage transitions by encoding conditional probabilities using a dynamic Bayesian network
Benchmarked performance against state-of-the-art neural network approaches using containerization (Docker) methods
Transcript in progress for journal publication by the end of 2025.
Role: Lead Architect & Author
Language: Python
Tools: CuPy, NumPy, PyEDF, OpenMP, PyTorch, TensorFlow
Dates: Aug 2024 - Present
Click here for Source Code

Performed Fourier Frequency Transformations on digital image signals to create feature detection algorithms for edges, shadows, and highlights
Applied in-painting operations
Role: Lead Developer
Language: Python
Tools: NumPy
Dates: Aug 2023 - Dec 2023
Click here for Source Code

Developed language modeling and speech synthesis techniques
Configured Large Language Model (LLM) parametrization to obtain the correct medical diagnosis for a symptomatic patient
Role: Lead Developer
Language: Python
Tools: NumPy
Dates: Aug 2023 - Dec 2023
Click here for Source Code

Designed and developed a text-based adventure game modeled for memory efficiency
Presented design
Role: Lead Program Designer, Lead Developer
Language: C++
Dates: Jan 2024 - May 2024
Click here to play the game! (coming soon)
Click here for Source Code

Developed algorithms to solve systems of linear equations using Gram-Schmidt Orthogonalization, Eigenfactorization, and Singular Value Decomposition
Determined convergence upon real roots of systems of non-linear equations using the Newton-Raphson Method and Gaussian elimination
Solved initial value problems & boundary value problems for elliptical, parabolic, and hyperbolic partial differential equations using first central & finite difference approximations
Role: Lead Developer
Language: Python
Tools: SciPy, NumPy
Dates: Jan 2023 - Aug 2024
Click here for Source Code

Developed algorithms using Bayesian Inference and Constraint Satisfaction Programming to guide a free agent using perceptions to search a grid space and perform a task while calculating probabilities of dangers en route
Role: Team Developer
Languages: Prolog, Python
Tools: Swiplserver
Dates: Jan 2024 - May 2024
Click here for Source Code

Developed a Long-Short Term Convolution Network Model (LSTM) using Hyperparameter Tuning and Optimization on a three-dimensional neural network
Data: body sensor signal data in a time series
Results: 20.5% loss with 91% consistent cross-validation accuracy
Role: Lead Developer
Language: Python
Tools: NumPy, Keras Tensorflow
Dates: May 2023 - Jul 2023
Click here for Source Code

Performed Big Data Analysis using data pipelines and machine learning techniques to classify tumors as cancerous
Data: medical image data of cancer tumors
Role: Lead Developer
Language: Python
Tools: Keras Tensorflow
Dates: Aug 2022 - Dec 2022
Click here for Source Code

Developed algorithm to perform forward propagation with backward chaining to find the optimal path between a search query and the objective destination using vectorization mechanisms
Implemented Constraint Satisfaction Programming
Role: Lead Developer
Language: Python
Tools: NumPy
Dates: Aug 2022 - Dec 2022
Click here for Source Code

Performed extensive preprocessing and data cleansing methods on a publicly-available data set notorious for missing data with consistent 71% classification accuracy, better than NumPy's original example online
Role: Lead Developer
Language: Python
Tools: Keras
Dates: Jan 2023 - Aug 2023
Click here for Source Code

Applied machine learning and statistical analysis to evaluate and report the predictions of misclassified recidivism rates using public data of exonerations and state government judicial systems
Publicly presented findings
Role: Team Developer, Presenter
Language: Python
Tools: Random Forest
Dates: Jan 2022 - May 2022
Click here for Source Code

Performed statistical data analysis to evaluate and report the World Health Organization's (WHO's) then-current world population data forecasts
Publicly presented findings
Role: Team Developer, Presenter
Language: R
Dates: Aug 2021 - Dec 2021
Click here for Source Code

Code for these and more projects are available at my GitHub.

Page updated

Google Sites

Report abuse