Projects
Course and personal projects. Updated as new work is completed.
Portfolio Website
CS 3870 · Secure Web Design
Completed 09/10/2025
Multi-page layout, responsive design, and deployed via GitHub Pages.
This portfolio website was built to showcase academic and personal projects. It features a clean multi-page layout, Bootstrap-based styling, and a polished responsive design.
Features
- Fully responsive using Bootstrap 5
- Custom styled components and polished UI
- Clean navigation with consistent page layout
- Project cards with expandable details
- Hosted using GitHub Pages
Tech Stack
HTML CSS Bootstrap 5 GitHub PagesRisk Management Plan
CYB 3930 · IT Security Management
CTRL ALT DEL
Solutions
Version 8.0 · Completed 5/18/2025
Comprehensive, multi-phase risk management plan covering IT infrastructure, assessments, mitigation, BIA/BCP, and full documentation aligned with NIST standards.
Overview
A complete enterprise-level Risk Management Plan for CTRL ALT DEL Solutions, covering infrastructure analysis, risk assessments, mitigation strategies, business impact analysis, and continuity planning.
Key Components
- IT Infrastructure Analysis (servers, AD, VPN, SIEM, cloud, firewalls)
- Risk Assessments using NIST 800-30
- Full Risk Register (likelihood, impact, severity)
- Risk Mitigation strategies for priority risks
- Business Impact Analysis (MAO, RPO, RTO)
- Business Continuity Plan development
Tools & Methods
- SimpleRisk platform for tracking assets and risks
- Gantt scheduling for assessment phases
- Asset valuation & threat modeling
- Compliance: NIST 800-171, NIST 800-37, NIST 800-34
Project Team
Sam Miller, Nick Thelen, Steven Weier
Hand-Written Digit Recognition
CS 4030 · Machine Learning
Completed 12/18/2024
A neural-network–based classifier trained on the UCI Optical Digit dataset (8×8 images). Explores model design, hyperparameter tuning, loss analysis, and performance evaluation using precision, recall, F1, confusion matrices, and DET curves.
Overview
A hand-crafted, fully connected neural network designed to classify 8×8 handwritten-digit images (0–9). Includes data processing, model design, training, and full performance validation.
Key Features
- Custom neural network implemented entirely in NumPy
- Hidden layer variations: 32, 64, 128 neurons (sigmoid activation)
- Softmax output layer for 10-class classification
- Trained using cross-entropy loss + backpropagation
- Early stopping with validation tracking
Dataset
- UCI Optical Recognition of Handwritten Digits
- Pre-processed 8×8 grayscale matrices (values 0–16)
- 5,620 samples (80% training, 20% validation/test)
Performance Evaluation
- Accuracy, precision, recall, and F1 score
- Confusion matrix & heatmap visualization
- One-vs-All DET curves for all classes
- Loss-trend comparison for each hidden-layer size
Tools Used
- NumPy, Matplotlib, Seaborn
- scikit-learn (metrics, splitting, encoding)
- Jupyter Notebook + Anaconda
Team Members
Brady Coenen · Zach Fisher · Hudson LePine · Nick Thelen
Project Software Implementation
CYB 3960 · Software Security
Completed 12/2/2024
Implementation of a secure task management system. Focuses on secure coding principles, data encryption, authentication, input validation, and integrity checks.
Implemented Security Features
- Strong authentication with complex passwords
- Data encryption at rest (user and task data)
- Data integrity using hashing (SHA-256) and digital signatures
- Buffer overflow prevention via Java's memory-safe mechanisms
- Backup options for user data
- Secure password storage with salted hashing
- Input validation and sanitation to prevent injection attacks
Unfinished Security Features
- Session timeout and management
- Account lockout mechanism after failed logins
- Race condition mitigation for concurrent operations
Project Focus
Implementing secure software according to the designed physical architecture while following best practices for authentication, encryption, input validation, and data integrity.
Team Members
Parker Stork · Pharis Martin · Nick Thelen