Yaochu Jin - Publications#
According to Google Scholar, his publications have received a total of over 26,000 citations and have an h-index of 76 as of June 20, 2021. This ranks him the 47th in the UK and 801th in the world in the discipline of Computer Science. He was named a "Highly Cited Researcher" by Web of Science in 2019 and 2020.
10 major publications:
[1] R. Cheng, Y. Jin, M. Olhofer and B. Sendhoff. A reference vector guided evolutionary algorithm for many-objective optimization. IEEE Transactions on Evolutionary Computation, 20(5):773 - 791, 2016 --
This paper, published in the top evolutionary computation journal (IF=11.169), proposes an evolutionary algorithm that scales well with the number of objectives and has become a popular evolutionary algorithm for solving problems with a large number of conflicting objectives. The algorithm was implemented in a software tool of Honda R&D Europe and successfully used for vehicle design. The nominee was the principle PhD supervisor of the first author, who received 2016 Surrey's Vice Chancellor's Postgraduate Study Award (annually one awardee only) and the 2018 "Outstanding PhD Dissertation Award" of the IEEE Computational Intelligence Society. This paper received over 580 citations and was an ESI hot paper.
[2] Y. Tian, R. Cheng, X. Zhang, Y. Jin. PlatEMO: A MATLAB platform for evolutionary multi-objective optimization. IEEE Computational Intelligence Magazine, 12(4): 73-87, 2017
Published in a flagship journal of the IEEE Computational Intelligence Society (IF=9.08), this paper reports a software tool for multi-objective evolutionary optimization, PlatEMO, which contains over 100 evolutionary algorithms, a large number of popular benchmark problems, as well as widely used performance indicators. It is increasingly popular among researchers in the evolutionary computation community. An adapted version of the tool was adopted by Honda R&D Europe in Germany for daily design optimization purposes. The paper won the 2019 "IEEE Computational Intelligence Magazine Outstanding Paper Award”. This paper has received over 450 citations.
[3] X. Zhang, Y. Tian, R. Cheng, Y. Jin. A decision variable clustering-based evolutionary algorithm for large-scale many-objective optimization. IEEE Transactions on Evolutionary Computation, 22(1):97-112, 2018.
Aiming to address large-scale optimization problems with many objectives, this is the first paper that is capable of handling problems having up to 5,000 decision variables and 10 objectives, whereas most state-of-the-art multi-objective evolutionary algorithms are verified on problems having less than 50 decision variables only. The paper is also published in the top journal of evolutionary computation, and received the 2020 "IEEE Transactions on Evolutionary Computation Outstanding Paper Award" of the IEEE Computational Intelligence Society (only one paper was awarded in 2020). This paper has received over 200 citations and is an ESI highly cited paper.
[4] Y. Tian, R. Cheng, X. Zhang, F. Cheng, Y. Jin. An indicator-based multi-objective evolutionary algorithm with reference point adaptation for better versatility. IEEE Transactions on Evolutionary Computation, 22 (4):609-622, 2017.
One big additional challenge in solving multi-objective optimization is that the shape of the Pareto front may vary from problem to problem, making many existing algorithms fail to work. By introducing a new performance indicator together with an adaptive reference point generation method, this paper develops a new algorithm for multi-objective optimization that performs effectively on a wide range of multi-objective optimization problems. This paper has received over 200 citations, and is an ESI highly cited paper.
[5] T. Chugh, Y. Jin, K. Miettinen, J. Hakanen, K. Sindhya. A surrogate-assisted reference vector guided evolutionary algorithm for computationally expensive many-objective optimization. IEEE Transactions on Evolutionary Computation, 21(1): 129-142, 2018.
Surrogate-assisted evolutionary algorithms aim to achieve a set of Pareto optimal solutions with a very limited computational budget for optimization of expensive problems. On the basis of the algorithm developed in [1]], this is the first paper that presents a surrogate-assisted evolutionary algorithm for solving expensive optimization problems with over 10 objectives. The algorithm has been successfully applied to a data-driven steel-making optimization problems, and a paper reporting its application to an air intake ventilation system received the "Best Student Paper Award" at the 2017 Congress on Evolutionary Computation. This paper has received 158 citations.
[6] X. Zhang, Y. Tian, R. Cheng and Y. Jin. An efficient approach to non-dominated sorting for evolutionary multi-objective optimization. IEEE Transactions on Evolutionary Computation, 19(2):201-213, 2015.
This paper proposes a computationally efficient method for non-dominated sorting by avoiding redundant comparisons of the dominance relationship between the solutions, significantly reducing the runtime of evolutionary algorithms for solving large-scale optimization problems. The proposed efficient sorting method can also be applied to other scientific problems where solutions need to be sorted according to multiple criteria. This paper has received over 320 citations, is an ESI highly cited paper, and won the 2017 "IEEE Transactions on Evolutionary Computation Outstanding Paper Award" (only one paper was awarded in 2017).
[7] H. Zhu and Y. Jin. Multi-objective evolutionary federated learning. IEEE Transactions on Neural Networks and Learning Systems, 31(4): 1310-1322, 2020.
This paper presents a distributed deep neural architecture search method based on an Erdos Rényi random graph and evolves both high-performance and communication-efficient neural network models. The developed algorithm is able to generate light-weight neural network models without sacrificing the performance. This work has attracted much interest from the neural network learning community, and the nominee was invited to give a Keynote Speech on federated learning at the 2019 International Symposium on Neural Networks (ISNN) and the 2020 Brazilian Conference on Intelligent Systems (BRACIS). The paper was published in a top journal of artificial intelligence (IF=8.793) and has received 58 citations.
[8] Y. Jin, M. Olhofer, B. Sendhoff. A framework for evolutionary optimization with approximate fitness functions. IEEE Transactions on Evolutionary Computation, 6 (5), 481-494, 2002.
This was the first paper that has proposed adaptive model management strategies to balance exploration and exploitation in surrogate-assisted evolutionary optimization. It also suggests a neural network training strategy that makes use of the search information based on a covariance matrix adaptation evolution strategy. The developed surrogate-assisted evolution strategy was also successfully applied to design optimization of a turbine blade. This is a pioneering work on surrogate-assisted evolutionary optimization and has received 660 citations. The nominee was invited to give an invited plenary speech on this topic at several international conferences, including one at 2016 World Congress on Computational Intelligence.
[9] Y. Jin and J. Branke. Evolutionary optimization in uncertain environments-a survey. IEEE Transactions on Evolutionary Computation, 9 (3), 303-317, 2005.
This is a seminal survey paper that has shaped a research field in evolutionary computation by establishing a unified framework for dynamic optimization, noisy optimization and robust optimization. Following the publication of the survey paper, several workshops, special sessions, and tutorials on the topic were organized by the nominee (some of them together with J. Branke). A task force on the same topic was also set up within the Evolutionary Computation Technical Committee of the IEEE Computational Intelligence Society (the nominee was the founding chair). Today, it has become a well-established research area in evolutionary computation. The paper has received over 1700 citations.
[10] Y. Jin. A comprehensive survey of fitness approximation in evolutionary computation. Soft computing, 9 (1):3-12, 2005.
Fitness approximation and surrogate-assisted evolutionary optimization were completely new to the majority of researchers in the evolutionary computation community twenty years ago. In fact, rather than a traditional survey paper, this is more a paper that set up a new research field by defining the problem, proposing basic ideas for problem solving, discussing main research challenges, and suggesting several lines of future research. Tutorials and workshops were organized by the nominee during the early 2000's at the Congress on Evolutionary Computation and Genetic and Evolutionary Computation Conference, two major conferences of evolutionary computation. This paper received over 1200 citations.
