Stephen Sommer

Who Am I?

AI/ML researcher & developer deeply curious about consciousness, machine learning, and the future of intelligence. I build intelligent systems, reflect on their implications, and dream of open-source breakthroughs that shape tomorrow.

Currently earning my Master's of Science in Computer Science โ€“ Machine Learning Specialization at Georgia Institute of Technology. I work full-time as a Software Developer at Fastenal, where I apply data engineering and machine learning to solve real-world problems. Based in Minnesota, I code with curiosity, hike deer trails with wonder, and practice dark arts rituals in the realm of neural networks.

Skills

Projects & Publications

๐Ÿ“„ Poster Presentation โ€“ Winona Research Celebration 2024

I presented โ€œEnhancing R2L Intrusion Detection Using Decision Treesโ€ at the Winona State University Research & Creative Achievement Celebration. This research investigates the application of Decision Tree machine learning algorithms to improve Remote-to-Local (R2L) intrusion detection in cloud computing environments, addressing critical concerns of data security and privacy in the age of AI and big data.

Utilizing the KDD Cup 1999 and NSL-KDD datasets, the study evaluates the performance of the algorithm on each dataset independently and on a combined dataset. Results show that combining the datasets significantly enhances detection accuracy, achieving 99.93% accuracy, 91% detection rate, and a false alarm rate of 0.08%. This outperforms the individual dataset results: KDD Cup 1999 with 99.90% accuracy, 83% detection rate, and 1.08% false alarm rate, and NSL-KDD with 99.83% accuracy, 83% detection rate, and 0.08% false alarm rate.

The findings demonstrate how machine learning, specifically decision trees, can adaptively and effectively safeguard data integrity by rapidly identifying suspicious activities within intrusion detection systems.

View Poster & Abstract

๐Ÿ› ๏ธ Featured Open-Source Projects

  • ๐Ÿ“Š statistical-array-analysis-python: Python utility for generating synthetic integer arrays and computing key statistical metrics using only standard libraries.
  • ๐Ÿ›ก๏ธ intrusion-detection-system-using-decision-trees: Jupyter Notebook demonstrating decision tree based machine learning for enhancing cybersecurity via R2L intrusion detection.
  • ๐ŸŽ™๏ธ AIPersonalAssistant: Modular Python voice assistant sandbox exploring speech interfaces, LLM integration, and human-AI interaction components.

See more on my GitHub profile.

๐Ÿ“š Upcoming Publications

  • Integrating symbolic and connectionist AI architectures
  • Philosophical analysis of AI alignment and sentience
  • Bridging stoic ethics and reinforcement learning models

Talk to My AI Assistant

This assistant is trained on AI research, neuroscience, machine learning, and philosophy of mind. Ask anything from beginner concepts to advanced topics!

Let's Connect