Tyler Streeter

RESEARCH FOCUS

To maximize my positive impact on society, I work toward simple, general machine learning principles, models, and algorithms which lead to useful applications and a foundation for safe, effective, human-level AI.

Specifically, I am driven to discover, through theory and experiment, machine learning methods that:

• have simple, powerful mathematical foundations
• encompass supervised, unsupervised, & reinforcement learning
• apply to both episodic environments and unending data sequences
• scale to high-dimensional data
• automatically avoid overfitting on small data sets
• are easy to evaluate and compare
• require few implementation decisions
• invoke a strong sense of confidence
• are easy to explain and implement
• are obvious in hindsight

My ideal method would effectively represent and learn structure from any environment encountered by the mammalian brain, thus enabling immediate practical applications among all common data domains. These include text, audio, images, video, stocks, 3D geometry, card/board games, video games, and robotics.

Most recently, I have focused on deriving information theoretic priors for Bayesian inference with undirected graph-based models. With this work I hope to provide a principled default solution to the problem of overfitting on small data sets.

EXPERTISE

EDUCATION

PhD Human Computer Interaction
ABD, Iowa State University, GPA 3.85/4.0
Thesis: Multivariate Binary Data Prediction with Log-Linear Models, Bayesian Inference, and Information Theoretic Priors.

MS Human Computer Interaction
2005, Iowa State University, GPA 3.92/4.0
Thesis: Design and Implementation of General Purpose Reinforcement Learning Agents.

BS Computer Engineering, Music Minor
2004, Iowa State University, GPA 3.83/4.0

EXPERIENCE

Independent Researcher (PhD Thesis) Oct 2019 - present
HCI Graduate Program, VR Applications Center
Iowa State University, Ames IA
Topics: Boltzmann machines, derivation of Bayesian priors with information theoretic foundations, inference and learning algorithms

Founder & Research Scientist, Brainpower Labs LLC Oct 2008 - present
Purpose: AI/ML research, software development to fund research.
Products: iBonsai for iOS, a meditative interactive 3D tree simulation (40,000 downloads in first 2 weeks).

Independent Researcher Aug 2006 - Dec 2013
HCI Graduate Program, VR Applications Center
Iowa State University, Ames IA
Topic: brain-inspired artificial general intelligence

Research Intern May 2006 - Aug 2006
Biometaphorical Computing Research Group
IBM TJ Watson Research Center, Yorktown Heights NY
Tasks: Implemented a novel cerebellum model, and tested motor learning on a simulated arm.

Teaching Assistant Jan 2006 - May 2006
Computational Perception course
Iowa State University, Ames IA
Topics: machine perception techniques and tools (e.g. OpenCV)

Independent Researcher (MS Thesis) Oct 2003 - Dec 2005
HCI Graduate Program, VR Applications Center
Iowa State University, Ames IA
Topics: reinforcement learning, artificial curiosity, planning, 3D physics simulation, interactive real-time visualization

Research Assistant May 2003 - Dec 2005
VR Applications Center
Iowa State University, Ames IA
Tasks: Developed VR software tools and applications for several funded research projects.

Research Assistant Aug 2002 - Dec 2003
Center for Nondestructive Evaluation
Iowa State University, Ames IA
Tasks: Assisted design and implementation of 3D software for ultrasonic probe simulations used for testing jet and space shuttle parts. Integrated STL file importer and octree space partitioning.

Research Assistant Jun 2002 - Aug 2002
Scalable Computing Lab
Ames Laboratory, Ames IA
Tasks: Developed general message logging software for computational chemistry and other applications on high performance clusters.

PRESENTATIONS

2019-10. Multivariate Binary Data Prediction with Log-Linear Models, Bayesian Inference, and Information Theoretic Priors. Iowa State University, Ames IA. Proposal presentation. (Not yet public.)

2009-07. Artificial Evolution. ISU Research Experience for Undergraduates. Presentation and demo. (Python demo of 2D catapult simulation and parameter evolution.)

2009-06. Sapience: A Brain-Inspired Cognitive Architecture. Iowa State University, Ames IA. Proposal presentation.

2009-03. A Hierarchical Empirical Bayesian Model of Cerebral Cortex. The 2nd Conference on Artificial General Intelligence, Arlington VA. Poster and demo. (Presented again in 2009-03 at Emerging Technologies Conference, Ames IA.)

2008-11. Artificial Evolution of Catapults and Neuroevolution. ISU Robotics Club, Ames IA. Presentation and demos. (Python demos of 2D catapult simulation and parameter evolution and 2D mobile robot neuroevolution with collision-based mating.)

2008-06. Artificial Evolution and Neuroevolution. ISU Research Experience for Undergraduates. Presentation and demo. (Python demo of 2D mobile robot neuroevolution with collision-based mating.)

2008-04. Brain-Inspired Artificial General Intelligence. Emerging Technologies Conference, Ames IA. Poster.

2006-11. Neuroevolution, Reinforcement Learning, Artificial Curiosity, and Self-Organizing Maps. ISU Robotics Club, Ames IA. Research presentation. (Presented again in 2006-12 to the ISU HCI Club.)

2006-10. An Integrated Cerebellum Model Explaining Associative Learning, Timing Prediction, and Motor Control. Society for Neuroscience Annual Meeting, Atlanta, GA. Poster and abstract. (Did not attend.)

2006-09. Verve: A General Purpose Open Source Reinforcement Learning Toolkit. ASME IDETC & CIE, Philadelphia PA. Paper presentation.

2006-08. A Model of Motor Task Offloading in the Cerebellum. IBM TJ Watson Research Center, Yorktown Heights NY. Poster.

2006-07. Design and Implementation of General Purpose Reinforcement Learning Agents. IBM TJ Watson Research Center, Yorktown Heights NY. MS thesis presentation. (Given to the Biometaphorical Computing Research Group.)

2006-07. Curiosity-Driven Exploration with Planning Trajectories. Twenty-First National Conference on Artificial Intelligence, Boston MA. Poster.

2006-04. Artificial Curiosity. HCI Forum, Ames IA. Research presentation.

2006-04. Artificial Curiosity. HCI Forum, Ames IA. Poster and demo.

2006-03. Design and Implementation of General Purpose Reinforcement Learning Agents. ISU Robotics Club, Ames IA. MS thesis presentation.

2005-11. Design and Implementation of General Purpose Reinforcement Learning Agents. Iowa State University, Ames IA. MS thesis defense.

2005-04. Motor Learning for Simulated Creatures. HCI Forum, Ames IA. Poster.

2005-04. OPAL: Open Physics Abstraction Layer. HCI Forum, Ames IA. Poster and demo.

2005-02. Neuroevolution and Reinforcement Learning for Physically Realistic Motor Control in Video Games. ISU Game Developers Club, Ames IA. Research presentation.

2005-01. Verve Research Overview. ISU Robotics Club, Ames IA. Research presentation.

2004-04. Autonomous Virtual Humans. HCI Forum and 8th International Immersive Projection Technology Workshop, Ames IA. Research presentation.

2003-11. Open Dynamics Engine and Simulated Humans with Neural Network Motor Controllers. ISU Game Developers Club, Ames IA. Tutorial and demos. (C++ demos of simulated humans with neural network motor controllers.)

(See tyl.st/artifacts for pdfs.)