Learning to Optimize Black-Box Functions with Extreme Limits on the Number of Function Evaluations

Ansótegui, Carlos; Sellmann, Meinolf; Shah, Tapan; Tierney, Kevin

doi:10.1007/978-3-030-92121-7_2

Cited by 4 publications

(4 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Across the board, the post hoc method works well. This is likely due our extremely small number of evaluations for algorithm configuration [25], leading to more benefits for exploration. The online method typically performs poorly in this setting and it is possible that more sophisticated bandit algorithms [58] could perform better.…”

Section: Discussionmentioning

confidence: 99%

“…Since robotics applications and datasets are reasonably expensive operations (compared to purely synthetic tasks), our work is related to work on extremely few function evaluations, typically on the order of 100 [25]. Typically this is studied in the Machine Learning community as hyperparameter search, finding configurations for ideal neural network configuration and training [26], [18].…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Direct Fitting of Gaussian Mixture Models

Keselman

Hebert

2019

2019 16th Conference on Computer and Robot Vision (CRV)

View full text Add to dashboard Cite

When fitting Gaussian Mixture Models to 3D geometry, the model is typically fit to point clouds, even when the shapes were obtained as 3D meshes. Here we present a formulation for fitting Gaussian Mixture Models (GMMs) directly to a triangular mesh instead of using points sampled from its surface. Part of this work analyzes a general formulation for evaluating likelihood of geometric objects. This modification enables fitting higher-quality GMMs under a wider range of initialization conditions. Additionally, models obtained from this fitting method are shown to produce an improvement in 3D registration for both meshes and RGB-D frames. This result is general and applicable to arbitrary geometric objects, including representing uncertainty from sensor measurements.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Direct Fitting of Gaussian Mixture Models

Keselman

Hebert

2019

2019 16th Conference on Computer and Robot Vision (CRV)

View full text Add to dashboard Cite

show abstract

“…We then assessed the accuracy of the training phase using the validation set. A Bayesian algorithm [25] was chosen for hyperparameter tuning. To tune the model, we adjusted the number of trees, feature fraction, learning rate, and the maximum depth of the trees.…”

Section: Methods Detailsmentioning

confidence: 99%

The fungal intestinal microbiota predict the development of bronchopulmonary dysplasia in very low birthweight newborns

Willis

Silverberg

Martin

et al. 2023

Preprint

View full text Add to dashboard Cite

Rationale: Bronchopulmonary dysplasia (BPD) is the most common morbidity affecting very preterm infants. Gut microbial communities contribute to multiple lung diseases, and alterations of the gut microbiome may be a factor in BPD pathogenesis. Objective: To determine if features of the multikingdom gut microbiome predict the development of BPD in very low birthweight newborns. Methods: We performed a prospective, observational cohort study comparing the multikingdom fecal microbiota of 147 very low birthweight infants with BPD or post-prematurity respiratory disease (PPRD) by sequencing bacterial 16S and fungal ITS2 ribosomal RNA. To address the potential causative relationship between gut dysbiosis and BPD, we used fecal microbiota transplant in an antibiotic-pseudohumanized mouse model. Comparisons were made using RNA sequencing, quantitative confocal microscopy, lung morphometry, and oscillometry. Measurements and Main Results: We analyzed 100 fecal microbiome samples collected during the second week of life. Infants who later developed BPD showed an obvious fungal dysbiosis as compared to infants with PPRD (P = 0.0209). Instead of fungal communities dominated by Candida and Saccharomyces, the microbiota of infants who developed BPD were characterized by the abundance of Aureobasidium and a greater diversity of rarer fungi in less interconnected community architectures. On successful colonization, the gut microbiota from infants with BPD augmented lung injury in the offspring of recipient animals. We identified alterations in the murine lung and intestinal microbiomes and transcriptional alterations associated with augmented lung injury. Conclusions: The gut fungal microbiome of infants who will develop BPD is dysbiotic and may contribute to disease pathogenesis.

show abstract

“…To run Differential Evolution algorithms on the BBOB-functions, we make use of the pyade package 25 , which is a python-based implementation used in the field (Ans ótegui et al, 2021;Nieto et al, 2021) that incorporates several variants of DE, including SHADE and L-SHADE employed for this study -see Section 2.2 for their description. We made some minor modifications to the base-code of pyade:…”

Section: Setup For the Experimentation On The Bbob Suitementioning

confidence: 99%

The importance of being constrained: dealing with infeasible solutions in Differential Evolution and beyond

Kononova¹,

Vermetten²,

Caraffini³

et al. 2022

Preprint

View full text Add to dashboard Cite

We argue that results produced by a heuristic optimisation algorithm cannot be considered reproducible unless the algorithm fully specifies what should be done with solutions generated outside the domain, even in the case of simple box constraints. Currently, in the field of heuristic optimisation, such specification is rarely mentioned or investigated due to the assumed triviality or insignificance of this question. Here, we demonstrate that, at least in algorithms based on Differential Evolution, this choice induces notably different behaviours -in terms of performance, disruptiveness and population diversity. This is shown theoretically (where possible) for standard Differential Evolution in the absence of selection pressure and experimentally for the standard and state-of-the-art Differential Evolution variants on special test function f0 and BBOB benchmarking suite, respectively. Moreover, we demonstrate that the importance of this choice quickly grows with problem's dimensionality. Different Evolution is not at all special in this regard -there is no reason to presume that other heuristic optimisers are not equally affected by the aforementioned algorithmic choice. Thus, we urge the field of heuristic optimisation to formalise and adopt the idea of a new algorithmic component in heuristic optimisers, which we call here a strategy of dealing with infeasible solutions. This component needs to be consistently (a) specified in algorithmic descriptions to guarantee reproducibility of results, (b) studied to better understand its impact on algorithm's performance in a wider sense and (c) included in the (automatic) algorithmic design. All of these should be done even for problems with box constraints.

show abstract

Learning to Optimize Black-Box Functions with Extreme Limits on the Number of Function Evaluations

Cited by 4 publications

References 27 publications

Direct Fitting of Gaussian Mixture Models

Direct Fitting of Gaussian Mixture Models

The fungal intestinal microbiota predict the development of bronchopulmonary dysplasia in very low birthweight newborns

The importance of being constrained: dealing with infeasible solutions in Differential Evolution and beyond

Contact Info

Product

Resources

About