Home

Welcome to COLA Laboratory #

Computational Optimization for Learning and Adaptive Systems (COLA) laboratory is working in computational/artificial intelligence, multi-objective optimization and decision-making, operational research, machine learning and statistical modeling of complex systems. Applications focus on search-based software engineering, energy, biological science and medicine.

We are constantly looking for talented and passionate people to join our lab. Please refer to vacancies for potential opportunities.
I have been working on the update of this website, though it has been a bit slow.

Recent News (Year 2024) #

  • Our paper “On the Hyperparameter Loss Landscapes of Machine Learning Models: An Exploratory Study” has been accepted to KDD 2025. This work developed tools to systematically understand the black box of the highly complex loss landscapes of hyperparameter optimization. There are several works underway in our lab along this line. Very well done, Mingyu! October 18
  • My single author paper “A Survey of Multi-objective Evolutionary Algorithm Based on Decomposition: Past and Future” has been accepted to IEEE Transactions on Evolutionary Computation. I see this as my 15 years’ early-career summary in the evolutionary multi-objective optimization domain, where I got my fame there. My lab has been marching towards several new interesting grand challenges for creating a better world in future. Nevertheless, deep down, I am a computational optimization scientist no matter what other areas I will be. Finally, I would like to appreciate my students who helped me to prepare this long survey article, including Mingyu, An, and Peili. October 21

  • Our paper “An Interpretable RNA Foundation Model for Exploration Functional RNA Motifs in Plants” has been accepted to Nature Machine Intelligence. This is the world first RNA foundation model pre-trained for the plant kingdom. While we only has 35 million parameters, this model can already beat other much larger models. Very appreciate to our student Heng’s hard work as well as our fantastic collaborators Haopeng and Yiliang in the John Innes Centre. Without this highly multi-disciplinary collaboration, we cannot achieve this far. There are many more follow-on works coming soon! October 16

  • Our paper “Direct Preference-Based Evolutionary Multi-Objective Optimization with Dueling Bandits” has been accepted to NeruIPS 2024. This is a super cool paper that achieves the first theoretical results of dueling bandits in multi-objective optimization. In particular, we significantly reduce the query complexity. Congratulations to Tian and Shengbo! September 25

  • We have two papers been accepted in EMNLP 2024. One is about adversarial defense in large language models while the other is about using language model to predict RNA secondary structure. Both of them are very cool, especially the RNA one. Congratulations to Heng! September 19

  • I’m honored to be recognized in the list of top 2% of the world most-cited scientists by Stanford University. I have been in this list since 2020. Really appreciate the hard work from my students and collaborators. September 19

  • Our paper on “Solving Expensive Optimization Problems in Dynamic Environments with Meta-learning” has been accepted to IEEE Trans. on Cybernetics. Congratulations to Huang! The initial idea can be extended to a wide range of other scenarios, even beyond the Bayesian optimization niche. July 25

  • Our paper on “A many-objective evolutionary algorithm based on interaction force and hybrid optimization mechanism” has been accepted to Swarm and Evolutionary Computation. Congratulations to Lei! July 20

  • I have been officially promoted to a senior academic position, Readership (i.e., Associate Professor++) in the UK system. Appreciate to the great work done by my group and support from my collaborators and mentors. Look forward to the next level promotion soon. July 16

  • Our paper on “TransOPT: Transfer Optimization System for Black-box Optimization” has been accepted to CIKM 2024 Demo Track. Congratulations to Peili! Very nice piece of initial work on a system for studying and developing transfer learning methods for black-box optimization. July 16

  • Our paper on “Evolutionary Alternating Direction Method of Multipliers for Constrained Multi-Objective Optimization with Unknown Constraints” has been accepted to IEEE Trans. on Evolutionary Comptation. Congratulations to Shuang! Very nice piece of work on problems with unknown constraints. This is has rarely been discussed yet highly important problem in the real world. July 3

  • Congratulations to Jack and Fan whose first ever RNA design benchmark is accepted to the ICML 2024 AI4Science workshop! Very well done! June 17

  • Congratulations to Phoenix who secured a position in GSK, one of the biggest pharmaceutical companies in the world. He will be exploring the responsible AI in drug discovery. Very big congratulations to Phoenix’s achievements. This also sets our new journey towards AI4Science. June 4

  • I am honored to be invited as a Senior Program Committee of CIKM 2024. This is a recognition of our group’s effort on data engineering in the past three years. May 17
  • Our AutoML for science work is featured in a WeiChat public account article for public outreaches. I’m very exciting to look forward more interesting works along this line of research. April 28

  • Our paper on “Multi-Output Framework for Time-Series Forecasting in Smart Grid Meets Data Scarcity” has been accepted to IEEE Trans. on Industrial Informatics. Congratulations to Jiangjiao! Very nice piece of work on time-series forecasting from the perspective of Gaussian process modeling. April 26

  • Our paper on “Evolutionary Multi-Objective Optimization for Contextual Adversarial Example Generation” has been accepted to FSE 2024! Congratulations to Shasha! Very nice piece of work on adversarial attacks in language models for software engineering. April 14

  • Our first multi-disciplinary paper about AutoML for science, named “iM-Seeker: A Webserver for DNA I-motifs Prediction and Scoring via Automated Machine Learning” has been accepted to Nucleic Acids Research. Congratulations to Fan and our collaborator in Ding’s lab. We have many more in coming! April 11

  • Our paper on “Evolutionary Art Attack For Black-Box Adversarial Example Generation” has been accepted to IEEE Trans. Evolutionary Computation. Congratulations to Phoenix! April 2

  • Our paper on “A Knee Point Driven Evolutionary Algorithm for Multi-Objective Bilevel Optimization” has been accepted to IEEE Trans. Cybernetics. Congratulations to Jiaxin! March 8

  • I’m thrilled to be selected as in the new cohort of 51 Turing Fellows in the UK. The Alan Turing Institute is the national institute of AI and data science. Look forward to developing new collaboration in my second term as a Turing Fellow. March 1

  • I have been invited to join the Editorial Board of Evolutionary Computation Journal (ECJ), effective since February! ECJ is one of the most prestigious journals in the evolutionary computation field. It is my honor to promote the visibility and impact of ECJ. January 25

  • Our papers on “An Automated Few-Shot Learning for Time Series Forecasting in Smart Grid Under Data Scarcity” has been accepted to IEEE Trans. Artificial Intelligence. Congratulations to Jiangjiao! January 19

  • Our papers on “Modeling Aspect Sentiment Coherency via Local Sentiment Aggregation” has been accepted to EACL 2024. Congratulations to Heng! January 18

Previous news can be found from our news archive.

Contact #

Dr. Ke Li
Department of Computer Science
Innovation Centre A1-D
University of Exeter
Streatham Campus, North Park Road
Exeter, EX4 4QF, UK
E-mail: k.li AT exeter.ac.uk
Tel: +(44) 0139-272-4557