diochnos/research/publications
[ DBLP ]
[Google Scholar]
[OpenReview Profile]
[Semantic Scholar]
[ORCiD]
[Scopus]
[ACM]
[LinkedIn]
[Twitter]
Research Interests: I am primarily interested in the design and analysis of machine learning algorithms with rigorous guarantees in the contexts of supervised learning, semi-supervised learning, adversarial learning, imbalanced data, and randomized search heuristics.
• Investigating these topics I try to provide statistical and computational bounds, or computational hardness results.
• In the absence of formal guarantees, I try to provide sound empirical conclusions.
During the last couple of years my students and I are investigating semi-supervised learning, regularization methods, learning with streaming data, and related topics.
-
A Review of Pseudo-Labeling for Computer Vision,
Patrick Kage, Jay Rothenberger, Pavlos Andreadis, and Dimitrios I. Diochnos.
Accepted for publication at the Journal of Artificial Intelligence Research (JAIR).
Please note that this is currently a pre-print and we intend to make some small changes so that we can align better with the final remarks of the reviewers as well as to include one repository that we identified after the submission to the journal. -
Dimensionally Reduced Open-World Clustering: DROWCULA,
Erencem Özbey and Dimitrios I. Diochnos.
This is work done while Erencem was a visiting undergraduate student at OU during the fall of 2024 and worked in this topic. Paper accepted in Australasian Joint Conference on Artificial Intelligence (AJCAI), 2025. -
Meta Co-Training: Two Views are Better than One,
Jay C. Rothenberger and Dimitrios I. Diochnos.
European Conference on Artificial Intelligence (ECAI), 2025.
🥇 Meta Co-Training is a semi-supervised learning method that achieves State-of-the-Art performance on ImageNet-10% and does very well on ImageNet-1% too.A preliminary version was presented in the Eighteenth International Symposium on Artificial Intelligence and Mathematics (ISAIM), 2024, as well as in a special session of the Joint Mathematics Meetings of the American Mathematical Society (JMM/AMS), 2025.
-
AI2ES:
The NSF AI Institute for Research on Trustworthy AI for Weather, Climate, and Coastal Oceanography,
Amy McGovern, Imme Ebert-Uphoff, Elizabeth A. Barnes, Ann Bostrom, Mariana G. Cains, Phillip Davis, Julie L. Demuth, Dimitrios I. Diochnos, Andrew H. Fagg, Philippe Tissot, John K. Williams.
AI Magazine, Volume 45, Issue 1, 2024. -
ISAIM-2022: international symposium on
artificial intelligence and mathematics,
Dimitrios I. Diochnos, Martin Charles Golumbic, Frederick Hoffman.
Foreword written due to the organization of the International Symposium on Artificial Intelligence and Mathematics (ISAIM) 2022. Appeared in Annals of Mathematics and Artificial Intelligence (AMAI), Volume 92, pages 1–4, (2024). -
Perceptrons Under
Verifiable Random Data Corruption,
Jose E. Aguilar Escamilla and Dimitrios I. Diochnos.
International Conference on Machine Learning, Optimization, and Data Science (LOD), 2023.
Work done when Jose was an undergraduate McNair Scholar in Computer Science. -
Research Issues in Adversarially Robust Stream-Based Federated Learning,
Abinash Borah, Dimitrios I. Diochnos, Le Gruenwald, Elaheh Jafarigol, Egawati Panjei, and Theodore B. Trafalis.
International Conference on Optimization and Learning (OLA), 2022. -
Evolving Monotone Conjunctions in Regimes Beyond Proved Convergence,
Pantia-Marina Alchirch, Dimitrios I. Diochnos, and Katia Papakonstantinopoulou.
Twenty-Fifth European Conference on Genetic Programming (EuroGP), 2022. -
Wind Prediction under Random Data Corruption (Student Abstract),
Conner Flansburg and Dimitrios I. Diochnos,
Thirty-Sixth AAAI Conference on Artificial Intelligence (AAAI), 2022, in the Student Abstract and Poster program. -
On the Evolvability of Monotone Conjunctions with an Evolutionary Mutation Mechanism,
Dimitrios I. Diochnos,
Journal of Artificial Intelligence Research (JAIR), 2021. -
Learning Reliable Rules under Class Imbalance,
Dimitrios I. Diochnos and Theodore B. Trafalis,
SIAM International Conference on Data Mining (SDM), 2021.
Supplementary material containing omitted proofs and discussion. -
Lower Bounds for Adversarially Robust PAC Learning under Evasion and Hybrid Attacks,
Dimitrios I. Diochnos, Saeed Mahloujifar, and Mohammad Mahmoody,
Nineteenth IEEE International Conference on Machine Learning and Applications (ICMLA), 2020.
A preliminary version appeared in the Sixteenth International Symposium on Artificial Intelligence and Mathematics (ISAIM), 2020.
An extended version of the paper is available on arXiv.org under the reference number 1906.05815. -
Foreword to special issue for ISAIM 2018,
Dimitrios I. Diochnos, Jürgen Dix, and Guillermo Simari,
Annals of Mathematics and Artificial Intelligence (AMAI), 2020. -
Understanding the Semantic Content of Sparse Word Embeddings Using a Commonsense Knowledge Base,
Vanda Balogh, Gábor Berend, Dimitrios I. Diochnos, and György Turán,
Thirty-Fourth AAAI Conference on Artificial Intelligence (AAAI), 2020.
A slighlty condensed version also appeared in the NeurIPS Workshop on Knowledge Representation & Reasoning Meets Machine Learning (KR2ML), 2019. -
Learning under p-tampering poisoning attacks,
Saeed Mahloujifar, Dimitrios I. Diochnos, and Mohammad Mahmoody,
Annals of Mathematics and Artificial Intelligence (AMAI), 2019. -
Curse of Concentration in Robust Learning: Evasion and Poisoning Attacks from Concentration of Measure,
Saeed Mahloujifar, Dimitrios I. Diochnos, and Mohammad Mahmoody,
Thirty-Third AAAI Conference on Artificial Intelligence (AAAI), 2019.
A preliminary version appeared in the NeurIPS Workshop on Security in Machine Learning (SECML), 2018.
The full version of the paper is available on arXiv.org under the reference number 1809.03063.
-
Interpretability of Hungarian embedding spaces using a knowledge base,
Vanda Balogh, Gábor Berend, Dimitrios I. Diochnos, György Turán, and Richárd Farkas,
Fifteenth Conference on Hungarian Computational Linguistics (MSZNY), 2019. -
Adversarial Risk and Robustness: General Definitions and Implications for the Uniform Distribution,
Dimitrios I. Diochnos, Saeed Mahloujifar, and Mohammad Mahmoody,
Thirty-Second Conference on Neural Information Processing Systems (NeurIPS), 2018.
The full version of the paper is available on arXiv.org under the reference number 1810.12272. -
Learning under p-Tampering Attacks,
Saeed Mahloujifar, Dimitrios I. Diochnos, and Mohammad Mahmoody,
Twenty-Ninth International Conference on Algorithmic Learning Theory (ALT), 2018.
A preliminary version appeared in the Fifteenth International Symposium on Artificial Intelligence and Mathematics (ISAIM), 2018.
An extended version of the paper is available on arXiv.org under the reference number 1711.03707. -
SmartOrch: An Adaptive Orchestration System for Human-Machine Collectives,
Michael Rovatsos, Dimitrios I. Diochnos, Zhenyu Wen, Sofia Ceppi, and Pavlos Andreadis,
Thirty-Second ACM Symposium on Applied Computing (SAC), 2017. -
On the Evolution of Monotone Conjunctions: Drilling for Best Approximations,
Dimitrios I. Diochnos,
Twenty-Seventh International Conference on Algorithmic Learning Theory (ALT), 2016.
The original publication is available at link.springer.com. -
Programming Model Elements for Hybrid Collaborative Adaptive Systems,
Ognjen Šćekić, Tommaso Schiavinotto, Dimitrios I. Diochnos, Michael Rovatsos, Hong-Linh Truong, Iacopo Carreras, and Schahram Dustdar,
First IEEE International Conference on Collaboration and Internet Computing (CIC), 2015. -
SmartSociety — A Platform for Collaborative People-Machine Computation,
Ognjen Šćekić, Daniele Miorandi, Tommaso Schiavinotto, Dimitrios I. Diochnos, Alethia Hume, Ronald Chenu-Abente, Hong-Linh Truong, Michael Rovatsos, Iacopo Carreras, Schahram Dustdar, and Fausto Giunchiglia,
Eighth IEEE International Conference on Service Oriented Computing & Applications (SOCA), 2015. -
Agent Protocols for Social Computation,
Michael Rovatsos, Dimitrios I. Diochnos, and Matei Craciun,
Second International Workshop on Multiagent Foundations of Social Computing, 2015.
-
Commonsense knowledge bases and network analysis,
Tanya Berger-Wolf, Dimitrios I. Diochnos, András London, András Pluhár, Robert H. Sloan, and György Turán,
Eleventh International Symposium on Logical Formalizations of Commonsense Reasoning, 2013.
-
On Multiple-Instance Learning of Halfspaces,
Dimitrios I. Diochnos, Robert H. Sloan, and György Turán,
Information Processing Letters (IPL), 2012.
The most recent online version can be found here. -
Leveling-Up in Heroes of Might and Magic III,
Dimitrios I. Diochnos,
Fifth International Conference on FUN WITH ALGORITHMS (FUN), 2010.
The original publication is available at link.springer.com. -
On Evolvability: The Swapping Algorithm, Product Distributions, and Covariance,
Dimitrios I. Diochnos and György Turán,
Fifth Symposium on Stochastic Algorithms, Foundations and Applications (SAGA), 2009.
The original publication is available at link.springer.com. - On the asymptotic and practical complexity of solving bivariate systems over the reals,
Dimitrios I. Diochnos, Ioannis Z. Emiris, and Elias P. Tsigaridas,
Journal of Symbolic Computation (JSC), 2009.
The paper is also available on arXiv.org under the reference number 1203.1017. -
Επίλυση Αλγεβρικών Συστημάτων Μικρής Διάστασης στους Πραγματικούς,
Δημήτρης Διώχνος,
Επιλεγμένες Πτυχιακές και Διπλωματικές Εργασίες,
Ετήσιο Βιβλίο, Τμήμα Πληροφορικής και Τηλεπικοινωνιών,
Εθνικό και Καποδιστριακό Πανεπιστήμιο Αθηνών, 2008. -
On the Complexity of Real Solving Bivariate Systems,
Dimitrios I. Diochnos, Ioannis Z. Emiris, and Elias P. Tsigaridas,
International Symposium on Symbolic and Algebraic Computation (ISSAC), 2007.