Talks by Thomas G. Dietterich


Highlighted Presentations

These presentations give good overviews of the work in my lab. Other presentations below are more specialized.

AI in Open Worlds: A Progress Report August 8, 2024. IJCAI 2024 Research Excellence Lecture. PDF Slides .

Uncertainty Quantification in Machine Learning July 12, 2024. OxML Summer School on Representation Learning. PDF Slides. Slightly revised from my OxML presentation.

Integrating Machine Learning into Safety-Critical Systems, July 18, 2024, ValgAI (Valencian Graduate AI Program). PDF Slides.

What's Wrong with Large Language Models and What We Should Be Building Instead, April 19, 2024, Johns Hopkins Institute for Assured Autonomy, PDF Slides; Youtube video

Anomaly Detection for OOD and Novel Category Detection, Keynote Speech, International Conference on Machine Learning Applications (ICMLA-2021), December 13, 2021. PDF slides.

Reinforcement Learning Prediction Intervals with Guaranteed Fidelity, Presentation at DARPA CAML PI Meeting, November 4, 2021. PDF slides.

Toward High-Reliability Artificial Intelligence, Beijing AI Ethics and Development Forum, August 3, 2021, Beijing, PRC. PDF slides.

Toward automated quality control for hydro-meteorological weather station dataData Science Africa, Kampala, July 25, 2020. This is the best current overview of our TAHMO work. PDF slides.

Advances in Anomaly Detection. Seminar, CMU, April 14, 2020. This is a good summary of our work at OSU on anomaly detection over the past 10 years. PDF slides.

Discovering and Removing Exogenous State Variables and Rewards for Reinforcement Learning. ICML 2018, Stockholm, Sweden. PDF slides.


Selected Presentations in Reverse Chronological Order


What's Wrong with Large Language Models and What We Should Be Building Instead, January 5, 2024, ACM CODS/COMAD Conference, Bangalore PDF slides

What's Wrong with Large Language Models and What We Should Be Building Instead, July 4, 2023, ValGRAI Seminar PDF slides. YouTube Video.

(Some) Steps Toward Trustworthy Machine Learning, Seminar, IIIA, Barcelona, Spain, July 23, 2021. PDF slides.

Machine Learning Methods for Robust Artificial Intelligence. Three lectures at Deep Learn 2021 (Canary Islands)

A Representation Analysis of Image Anomaly Detection. CVPR 2021 Workshop on Open World Computer Vision. PDF slides.

Anomaly Detection in Machine Learning and Computer Vision. Seminar at Booking.Com. March 8, 2021. PDF slides.

What High-Reliability Human Organizations can Teach Us about Robust Artificial Intelligence, Decision Theory and the Future of Artificial Intelligence. August 26, 2019, Canberra, Australia. PDF Slides

Research Methods in Machine Learning, Keynote Speech, New in ML, NeurIPS-2019, December 9, 2019, Vancouver, BC. PDF slides

Robust Artificial Intelligence: Why and How, Keynote Speech, FLAIRS-2017, May 22, 2017. PDF slides.

Anomaly Detection: Principles, Benchmarking, Explanation, and Theory, Keynote Speech, ICML 2016 Workshop on Anomaly Detection, June 24, 2016. PDF slides.

Machine Learning for Sustainable Development and Biological Conservation, OSTP Workshop on AI For Social Good, Washington, DC, June 7, 2016. PDF slides.

Artificial Intelligence and Robotics at Oregon State University, Huawei Noah's Ark Lab, Hong Kong, May 15, 2016. PDF slides.

Steps Toward Robust Artificial Intelligences, AAAI President's Address. AAAI 2016. Phoenix, AX. February 14, 2016. PDF slides

Smart Software in a World with Risk, DARPA Wait, What?, September 20, 2015. Video.

Efficient Sampling for Simulator-Defined MDPs, European Workshop on Reinforcement Learning (EWRL-2015), July 10, 2015. PDF slides. Note: The equivalent trajectories algorithm presented in this talk is incorrect and underestimates the variance. A paper with a corrected algorithm is under review.

Advances in Anomaly Detection, Stanford Data Science Infoseminar, February 27, 2015. PDF slides.

Modeling bird migration by combining weather radar and citizen science data, Oberlin College, May 8, 2014. PDF slides.

Challenges for Machine Learning in Computational Sustainability, Columbia University Data Science Seminar, March 13, 2014. Video

Challenges for Machine Learning in Computational Sustainability, Posner Lecture, NIPS-2012, South Lake Tahoe, December 5, 2012. PDF slides. Video at videolectures.net.

Machine Learning for Ecological Science and Environmental Policy, Jornadas Chilenas de Computacion, Tutorial, Valparaiso, Chile, November 14, 2012. PDF slides.

Graphical Models and Flexible Classifiers: Bridging the Gap with Boosted Regression Trees, Jornadas Chilenas de Computacion, Keynote Speech, Valparaiso, Chile. November 12, 2012. PDF slides.

Machine Learning and Computational Sustainability, Keynote Speech, Brazilian Society for Neural Networks, Curitiba, Parana, Brazil. October 22, 2012. PDF slides.

Bridging the two cultures: Latent variable statistical modeling with boosted regression trees, Ecological Society of America (ESA-2012), Portland, Oregon, August 7, 2012. PDF slides.

Machine Learning Methods for Timing of Biological Events. Second Annual Workshop on Understanding Climate Change from Data, Minneapolis, Minnesota, August 6-7, 2012 PDF slides.

Computational Sustainability: Applying Advanced Computing to Ecological Science and Ecosystem Management, Invited Speech, 4th International Symposium on IT Convergence Engineering (ISITCE 2012), Seoul, S. Korea, July 13, 2012. PDF slides.

Machine Learning for Computational Sustainability, Invited Talk, Third International Green Computing Conference, San Jose, California, June 6, 2012. PDF slides.

Novel machine learning methods for learning models of bird distribution and migration from citizen science data. NICTA Canberra, May 15, 2012. PDF slides.

Challenges for Machine Learning in Ecological Science and Environmental Management. Research Triangle Distinguished Lecture Series, January 23, 2012. PDF slides.

Graphical Models and Flexible Classifiers: Bridging the Gap with Boosted Regression Trees. The 2011 Conference on Technologies and Applications of Artificial Intelligence (TAAI 2011), Keynote Speech, November 12, 2011. PDF slides.

Inferring moth emergence from abundance data: A novel mathematical approach using birth-death contingency tables. Ecological Society of America (ESA-2011), Austin, Texas, August 10, 2011. PDF slides.

Computer Vision Seminar, Caltech, January 19, 2011. PDF slides.

First Asian Conference on Machine Learning, Nanjing, China, November 1, 2009. PDF slides.

Machine Learning in Ecosystem Informatics, International Joint Conference on Artificial Intelligence, Pasadena, CA. July 16, 2009. PDF slides.

Activity Recognition in TaskTracer and CALO, Invited Talk, Workshop on Plan, Activity, and Intent Recognition, International Joint Conference on Artificial Intelligence, Pasadena, CA, July 11, 2009. PDF slides

Learning in an Integrated Intelligent System: Examples from the CALO System, Keynote Address, IBM Haifa Machine Learning Workshop, Haifa, Israel, May 25, 2008. PDF slides.

Machine Learning in Ecosystem Informatics, Keynote Address, Discovery Science 2007, October 1, 2007. PDF slides.

Hierarchical Reinforcement Learning, Air Force Research Laboratory, Rome, NY, September 21, 2006. PDF slides.

Three Challenges for Machine Learning Research, IBERAMIA-2004 (Iberian-American Artificial Intelligence Conference), Puebla, Mexico, November 24, 2004. PDF slides.

Fitting Conditional Random Fields via Gradient Boosting, Department of Computer and Information Sciences, University of Pennsylvania, April 21, 2003. PDF slides.

Sequential Supervised Learning: General Methods for Sequence Labeling and Segmentation. Invited talk, 2003 IEEE International Conference on Data Mining, Melbourne, Florida. PDF slides.

Low Bias Bagged Support Vector Machines. Giorgio Valentini and Thomas G. Dietterich. 2003 International Conference on Machine Learning. PDF slides.

Bias-Variance Analysis of Ensemble Learning. Lecture presented at the IIASS International School on Neural Nets "E.R. Caianiello" 7th Course, Ensemble Methods for Learning Machines. 23 September, 2002. Power point presentation.

Ensembles for Cost-Sensitive Learning. Lecture presented at the IIASS International School on Neural Nets "E.R. Caianiello" 7th Course, Ensemble Methods for Learning Machines. 25 September, 2002. Power point presentation.

Machine Learning for Sequential Data. Invited talk delivered to the Joint IAPR International Workshops on Syntactical and Structural Pattern Recognition (SSPR 2002) and Statistical Pattern Recognition (SPR 2002). Windsor, Canada, August 6-9, 2002. Power point presentation.

Machine Learning: Making Computer Science Scientific Forsythe Memorial Lecture, Stanford University, March 18, 2002. Power Point presentation.

Hierarchical Reinforcement Learning Tutorial given at The Sixteenth International Conference on Machine Learning Bled, Slovenia, June 27, 1999. Postscript slides.

Connectionist Supervised Learning: An Engineering Approach, Tutorial co-authored with Andreas S. Weigend and given at Eleventh Internation Conference on Machine Learning (ML-94), New Brunswick, NJ, July, 1994. Postscript slides.

Machine Learning: Issues, Answers, and Quandaries. Invited Speech, Ninth National Conference on Artificial Intelligence (AAAI-91), July, 1991. Postscript slides.