These projects represent individual work that I can share code for publicly. Each project focuses on building and reasoning about software systems, with an emphasis on correctness, structure, and real-world constraints.
Extended a provided Unix shell skeleton with process execution, signal handling, and persistent command history.
C · Unix · fork/exec · Signals
View Details GitHubPython simulation modeling demyelination and remyelination cycles using a graph-based neural representation.
Python · NetworkX · Simulation
View Details GitHubApplied regression modeling project focused on exploratory data analysis, model diagnostics, and statistically sound interpretation of real-world data.
View Project DetailsThis project focused on applying statistical modeling techniques to real-world data in order to understand relationships between variables, evaluate model assumptions, and draw defensible conclusions from noisy datasets. The work emphasized the full analysis pipeline, from exploratory data analysis through model refinement and interpretation.
I conducted exploratory data analysis to identify trends, outliers, and potential violations of linear regression assumptions. Multiple regression models were developed and compared using transformations, interaction terms, and diagnostic measures to improve model fit and explanatory power. Throughout the process, I evaluated residual behavior, multicollinearity, and variance structure to ensure that assumptions were reasonably satisfied.
Model performance was assessed using goodness-of-fit metrics and hypothesis testing, with an emphasis on interpreting coefficients in a meaningful and context-aware way. The final analysis focused on what the model could reliably explain, where uncertainty remained, and how data limitations could influence conclusions.
This project extends a provided Unix shell skeleton by implementing core systems-level functionality related to process execution, signal handling, and persistent command history. My work focused on integrating operating system concepts into an existing command-parsing framework rather than building a shell entirely from scratch.
I implemented process creation and execution using fork,
execvp, and wait, enabling both foreground and
background command execution. I also added a persistent command history
mechanism backed by file I/O, allowing commands to be stored across sessions
and re-executed through a custom history recall interface.
The shell supports asynchronous behavior through signal handling, including
a SIGINT handler that allows users to view command history without
terminating the shell. Internally, recent commands are managed using a
fixed-size circular buffer, requiring careful state management and edge-case
handling.
I built a Python-based simulation to model simplified cycles of demyelination and remyelination in Multiple Sclerosis. The simulation represents neurons as nodes in a randomly generated graph and tracks neuron health over time using configurable damage and recovery rates.
The system initializes a neural network using NetworkX, assigns health values to each neuron, and visualizes disease progression through color-coded nodes. The simulation alternates between demyelination and remyelination phases, where neurons probabilistically lose or regain health based on parameterized inputs.