Tyler Streeter

Summary

I am a researcher and software engineer with a long-standing interest in creating artificial general intelligence. Having spent years studying a wide range of machine learning models and methods, I now focus on deriving new learning algorithms from simple, elegant principles.

I enjoy writing clean code and simple APIs, designing data visualizations to gain intuition about new domains, simulating physical processes with unexpected emergent behavior, building tangible objects from humble materials, and capturing big ideas with small math. My ideal project is one that lets me be a scientist, artist, and engineer.

Highlights

• Machine Learning Researcher, Engineer, Founder at Brainpower Labs since 2008
• PhD Machine Learning (2021), MS Reinforcement Learning, BS Computer Engineering
• Expertise:
  • C/C++, Python, 3D simulation, visualization, parallelization, open source
  • machine learning, graphical models, neural networks, optimization, MCMC, reinforcement learning, information theory, Bayesian methods, brain-inspired AI
• Research projects:
  • Theory and learning algorithms for binary log-linear models with objective priors.
  • Brain-inspired architecture for artificial general intelligence.
  • Infomax mixture models with hierarchy and topographic constraints.
  • Cerebellum computational model at IBM Research.
  • Reinforcement learning with various state spaces, planning, and artificial curiosity.
  • Evolution of motor control in 3D simulated humanoids.
• Software projects:
  • Sample Lab: visualization tool for graph-based Monte Carlo algorithms in C.
  • Sapience: AGI library in C++ and Python, debugger, and test environments.
  • Verve: general purpose reinforcement learning in C++ and Python.
  • OPAL: Open Physics Abstraction Layer, simple 3D physics simulation in C++.
  • Voce: tiny API for speech synthesis and recognition in C++ and Java.
  • QuickMP: simple loop parallelization macros in a single C++ header file.
  • QuickProf: simple performance profiling in a single C++ header file.
  • QuickTest: simple unit testing in a single C++ header file.
  • iBonsai: meditative interactive 3D tree simulation in C++ for iOS.

Education

Experience

Open Source Software

Commercial Software

Proprietary Software

Publications & Written Work

Presentations

Honors

• Robbins Graduate Scholarship
• National Merit Finalist
• ISU College of Engineering Scholarships
• Dean's List Every Semester
• Undergraduate Research Assistantship
• Student Finalist (highest award), Independent Games Festival
• Lead Tenor Saxophone, ISU Jazz Ensemble

Volunteer Experience

• Conference Associate (4 years), Game Developers Conference
• Question Reviewer/Scorekeeper (2 years), US DOE Science Bowl
• Mentor, NSF Research Experience for Undergraduates
• Technical Judge, FIRST Lego League Iowa
• Fundraiser/Laborer, humanitarian trips to six countries

Software Development Expertise

• languages: C/C++, Python, LaTeX, Mathematica, HTML
• libraries: SDL, ODE, NumPy, SciPy, Matplotlib
• OS targets: macOS, Linux, Windows, iOS
• programs: Vim, CMake, GIMP, Inkscape, Blender, Audacity
• parallelization: multithreading, SIMD, OpenMP, MPI
• visualization: 2D and 3D graphics, OpenGL, ray tracing, VR
• simulation: collision detection, rigid body dynamics, embodied AI simulations (humanoids/bipeds, arms/hands, wheeled robots)
• miscellaneous: clean code, good documentation, open source, cross platform development, AI/ML visualization tool development, video game architectures, cellular automata, cell-based physics models, speech recognition, speech synthesis, web development

Machine Learning Expertise

• undirected graphical models: Boltzmann machines, Markov random fields, Ising models, spin glasses, Hopfield networks, energy-based models, log-linear models
• directed graphical models: feedforward neural networks, Bayesian networks, logistic belief networks
• other models: self-organizing maps, generative models, kernel mixture models
• optimization algorithms: gradient descent, coordinate descent, 2nd-order methods, natural gradient, simulated annealing, genetic and evolutionary algorithms, deep learning
• MCMC sampling algorithms: Gibbs, Metropolis, exact/perfect sampling, partition function/free energy estimation
• learning objectives: supervised learning, unsupervised learning, reinforcement learning, clustering, lossless data compression, text/sequence prediction
• general principles: information theory, Bayesian methods, objective priors, maximum entropy, minimum description length (MDL), Fisher information, computational neuroscience, brain-inspired AI, artificial curiosity

Favorite Textbooks

• MacKay. Information Theory, Inference, and Learning Algorithms.
• Sutton & Barto. Reinforcement Learning: An Introduction.
• Russell & Norvig. Artificial Intelligence: A Modern Approach.
• Koller & Friedman. Probabilistic Graphical Models.
• Haykin. Neural Networks and Learning Machines.

Relevant Coursework

• general: Calculus, Differential Equations, Linear Algebra & Matrices, Discrete Math, Classical Physics, General Chemistry, Electric Circuits, Electronic Devices, Vehicle Dynamics
• computer software: C Programming, C++ Programming, Data Structures, Algorithm Design and Analysis, Software Engineering, Operating Systems, Networking, Microcontrollers & Assembly Languages, High Performance Computing, Computer Architecture
• computer graphics: Computer Animation and Modeling, Computer Graphics, Advanced Computer Graphics, Game Design and Development, Virtual Reality
• AI & machine learning: Artificial Intelligence, Statistics, Advanced Bayesian Methods, Probabilistic Graphical Models, Computational Perception, Developmental Robotics, Complex Adaptive Systems, Evolutionary Computation, iCub Summer School
• neuroscience: Brain and Behavior, Neural Basis of Movement, Neurobiology
• human computer interaction: Cognitive Psychology of HCI, Topics in HCI, Interaction Techniques

Other Interests

• general: running, drawing, cooking, piano, saxophone, video games (especially NES, SNES, DOS), making things from scratch
• tangible: bookbinding, woodworking, prototyping, paper electronics, cardboard robotics, artificial muscles, computer keyboards from scratch, molding & casting techniques, adhesives/coatings/binders, making everyday things from common materials
• mental: mathematics, physics, computer science, unix, handwritten notes, hand-drawn diagrams, old textbooks, math typesetting, finding better ways to represent & organize ideas (both paper & digital), brainstorming new projects & startups, thinking big, minimizing description lengths, discretizing continua, getting to the bottom of things