Biasing Bayesian Optimization Algorithm using Case Based Reasoning

Kaedi, Marjan; Ghasem-Aghaee, Nasser

doi:10.1016/j.knosys.2011.05.015

Cited by 21 publications

(14 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The latter is widely applied nowadays, particularly in areas dealing with the classification and optimisation problems [17][18][19][20][21]. For the purposes of this paper, the performance of the methodology was examined by testing the dynamic stability of generators and statistically analysing test results.…”

Section: Introductionmentioning

confidence: 99%

Application of Bayesian method for electrical power system transient stability assessment

Augutis

Žutautaitė

Radziukynas

et al. 2012

International Journal of Electrical Power & Energy Systems

View full text Add to dashboard Cite

Section: Introductionmentioning

confidence: 99%

Application of Bayesian method for electrical power system transient stability assessment

Augutis

Žutautaitė

Radziukynas

et al. 2012

International Journal of Electrical Power & Energy Systems

View full text Add to dashboard Cite

“…Over the years, a handful of research advances for global optimization under the theme of transfer optimization generally fall into the following two categories: (1) store a pool of solutions from previous problems that can be subsequently reused to guide Chapter 2. Literature Review the search on the optimization problem of interest [11,14]; (2) directly reuse the (probabilistic) models that are built during past problem-solving exercises [15,27]. The proposed approaches in the present thesis also fall into these two categories.…”

Section: Chapter Literature Reviewmentioning

confidence: 99%

“…Over recent years, there have been a handful of methods initiating cross-task learning in optimization. Different modes of knowledge transfer have been proposed in this regard, including the direct injection of raw solutions [11,14,16], biasing search through prior solution distribution models (that provide useful hints on where to search on a related task [15,27]), and transfer regression in surrogate-assisted optimization [13,32]. While positive results have been reported in various domains, for e.g., in operations research, engineering optimization, neuro controller design, etc., it is found that the learning capabilities of methods of the aforementioned type are typically limited by the relative simplicity of the machine learning models used.…”

Section: Predicting the Optimum For Combinatorial Optimization Problemsmentioning

confidence: 99%

“…In addition to direct genetic transfer from source to target evolutionary searches [40], higher-order model-based transfer optimization algorithms are also explored in the literature. In [27], a heuristic criterion is used within a specific class of combinatorial problems for retrieving related source probabilistic models from a database to bias the target optimization search. In [14,41], the model takes the form of a positive semidefinite (PSD) matrix that induces a modified (biased) distance metric for graph-based clustering cum sequencing problems.…”

Section: Online Transfer Optimization Techniquesmentioning

confidence: 99%

See 1 more Smart Citation

Methods in multi-source data-driven transfer optimization

Da¹

View full text Add to dashboard Cite

show abstract

“…Over the years, a handful of success stories have surfaced in the relatively young field of transfer optimization (TO), encompassing sequential transfers [14][15][16][17][18][19] as well as multitasking [20][21][22]. Similar to transfer learning, sequential transfer optimization utilizes high quality solutions from various source problems to solve a target optimization task.…”

Section: And Profmentioning

confidence: 99%

En route to automated extraction and transfer of knowledge in multitask optimization : an evolutionary perspective

Bali¹

View full text Add to dashboard Cite

It is a conventional wisdom that real world problems seldom occur in isolation. The motivation for this work, inspired from the observation that humans rarely tackle every problem from scratch, is to improve optimization performance through adaptive knowledge transfer across related problems. The scope for spontaneous transfers under the simultaneous occurrence of multiple problems unveils the benefits of multitasking. Multitask optimization has recently demonstrated competence in solving multiple (related) optimization tasks concurrently. Notably, in the presence of underlying relationships between problems, the transfer of high quality solutions across them has shown to facilitate superior performance characteristics -as the cost of re-exploring overlapping regions of the search space is reduced. However, in the absence of any prior knowledge about the inter-task synergies (as is often the case with general black-box optimization), the threat of predominantly negative transfer prevails. Susceptibility to negative inter-task interactions can in fact be detrimental, often impeding the overall convergence behavior. To allay such fears, this thesis presents viable solutions towards automated extraction and transfer of (fruitful) knowledge such that any deleterious e↵ects of otherwise negative intertask exchanges are suppressed. To this end, an in-depth theoretical analysis is first conducted to unveil the primary caveats which concern the global convergence characteristics of the present day multitasking evolutionary optimization framework.Next, a novel evolutionary computation framework is proposed that enables online learning and exploitation of the similarities (and discrepancies) between distinct tasks in multitask settings via probabilistic mixture models. The proposed method is based on principled theoretical arguments that seek to minimize the tendency of harmful interactions between tasks, based on a purely data-driven learning of relationships among them. As a proof of concept, the method is initially validated experimentally on a wide range of synthetic discrete and continuous single-objective benchmarks. Thereafter, a realization of similar concepts of the proposed method is extended to the domain of multi-objective optimization (an omnipresent scenario in our daily lives). It is noteworthy that this work shall be among the first to utilize xiii xiv probabilistic modeling to capture inter-task relationships between multi-objective optimization tasks in the context of evolutionary multitasking. Empirical studies on a series of benchmark test functions show that the method is able to decipher and adapt to the degree of similarity between distinct multi-objective optimization tasks on the fly. Finally, the practicality of the proposed methods are substantiated on various real world case studies including reinforcement learning, multi-fidelity optimization and evolutionary deep learning. Not only do the practical studies provide insights into the behavior of the methods in the face of several (many) complex ...

show abstract

Biasing Bayesian Optimization Algorithm using Case Based Reasoning

Cited by 21 publications

References 28 publications

Application of Bayesian method for electrical power system transient stability assessment

Application of Bayesian method for electrical power system transient stability assessment

Methods in multi-source data-driven transfer optimization

En route to automated extraction and transfer of knowledge in multitask optimization : an evolutionary perspective

Contact Info

Product

Resources

About