{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,11]],"date-time":"2026-07-11T05:48:20Z","timestamp":1783748900535,"version":"3.55.0"},"reference-count":35,"publisher":"Springer Science and Business Media LLC","issue":"8","license":[{"start":{"date-parts":[[2023,5,22]],"date-time":"2023-05-22T00:00:00Z","timestamp":1684713600000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2023,5,22]],"date-time":"2023-05-22T00:00:00Z","timestamp":1684713600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"DOI":"10.13039\/501100002954","name":"Universit\u00e0 degli Studi di Milano - Bicocca","doi-asserted-by":"crossref","id":[{"id":"10.13039\/501100002954","id-type":"DOI","asserted-by":"crossref"}]}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["J Ambient Intell Human Comput"],"published-print":{"date-parts":[[2023,8]]},"abstract":"<jats:title>Abstract<\/jats:title><jats:p>Bayesian optimization (BO) based on the Gaussian process model (GP-BO) has become the most used approach for the global optimization of black-box functions and computationally expensive optimization problems. BO has proved its sample efficiency and its versatility in a wide range of engineering and machine learning problems. A limiting factor in its applications is the difficulty of scaling over 15\u201320 dimensions. In order to mitigate this drawback, it has been remarked that optimization problems can have a lower intrinsic dimensionality. Several optimization strategies, built on this observation, map the original problem into a lower dimension manifold. In this paper we take a novel approach mapping the original problem into a space of discrete probability distributions endowed with a Wasserstein metric. The Wasserstein space is a non-linear manifold whose elements are discrete probability distributions. The input of the Gaussian process is given by discrete probability distributions and the acquisition function becomes a functional in the Wasserstein space. The minimizer of the acquisition functional in the Wasserstein space is then mapped back to the original space using a neural network. Computational results for three test functions with dimensionality ranging from 5 to 100, show that the exploration in the Wasserstein space is significantly more effective than that performed by plain Bayesian optimization in the Euclidean space and its advantage grows with the dimensions of the search space.<\/jats:p>","DOI":"10.1007\/s12652-023-04640-7","type":"journal-article","created":{"date-parts":[[2023,5,22]],"date-time":"2023-05-22T18:02:44Z","timestamp":1684778564000},"page":"11263-11271","update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":8,"title":["Wasserstein enabled Bayesian optimization of composite functions"],"prefix":"10.1007","volume":"14","author":[{"given":"Antonio","family":"Candelieri","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4187-4209","authenticated-orcid":false,"given":"Andrea","family":"Ponti","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Francesco","family":"Archetti","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2023,5,22]]},"reference":[{"key":"4640_CR1","first-page":"21428","volume":"33","author":"D Alvarez-Melis","year":"2020","unstructured":"Alvarez-Melis D, Fusi N (2020) Geometric dataset distances via optimal transport. Adv Neural Inf Process Syst 33:21428\u201321439","journal-title":"Adv Neural Inf Process Syst"},{"key":"4640_CR2","unstructured":"Alvarez-Melis D, Fusi N (2021) Dataset dynamics via gradient flows in probability space. In: International Conference on machine learning. PMLR, pp 219\u2013230"},{"key":"4640_CR3","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-24494-1","volume-title":"Bayesian optimization and data science","author":"F Archetti","year":"2019","unstructured":"Archetti F, Candelieri A (2019) Bayesian optimization and data science. Springer International Publishing"},{"key":"4640_CR34","unstructured":"Arjovsky M, Chintala S, Bottou L (2017) Wasserstein generative adversarial networks. In: International conference on machine learning. PMLR, pp 214-223"},{"key":"4640_CR4","first-page":"21524","volume":"33","author":"M Balandat","year":"2020","unstructured":"Balandat M, Karrer B, Jiang D, Daulton S, Letham B, Wilson AG, Bakshy E (2020) BoTorch: a framework for efficient Monte-Carlo Bayesian optimization. Adv Neural Inf Process Syst 33:21524\u201321538","journal-title":"Adv Neural Inf Process Syst"},{"key":"4640_CR5","doi-asserted-by":"crossref","unstructured":"Candelieri A, Ponti A, Archetti F (2021) Uncertainty quantification and exploration\u2013exploitation trade-off in humans. J Ambient Intell Humaniz Comput, 1\u201334.","DOI":"10.1007\/s12652-021-03547-5"},{"key":"4640_CR6","doi-asserted-by":"crossref","unstructured":"Candelieri A, Ponti A, Archetti F (2022a) Bayesian optimization in Wasserstein spaces. In: International Conference on Learning and Intelligent Optimization. Springer, Cham","DOI":"10.1007\/978-3-031-24866-5_19"},{"key":"4640_CR7","unstructured":"Candelieri A, Ponti A, Archetti F (2022b) Gaussian Process regression over discrete probability measures: on the non-stationarity relation between Euclidean and Wasserstein Squared Exponential Kernels. arXiv preprint arXiv:2212.01310"},{"key":"4640_CR8","unstructured":"Crovini E, Cotter SL, Zygalakis K, Duncan AB (2022) Batch Bayesian optimization via particle gradient Flows. arXiv preprint arXiv:2209.04722"},{"key":"4640_CR9","doi-asserted-by":"crossref","unstructured":"Frazier PI (2018) Bayesian optimization. In: Recent advances in optimization and modeling of contemporary problems. INFORMS, pp 255\u2013278","DOI":"10.1287\/educ.2018.0188"},{"key":"4640_CR10","unstructured":"Frogner C, Mirzazadeh F, Solomon J (2019) Learning embeddings into entropic Wasserstein spaces.\u00a0arXiv preprint arXiv:1905.03329"},{"key":"4640_CR11","doi-asserted-by":"publisher","first-page":"705","DOI":"10.1090\/qam\/1788425","volume":"58","author":"W Gangbo","year":"2000","unstructured":"Gangbo W, McCann RJ (2000) Shape recognition via Wasserstein distance. Q Appl Math 58:705\u2013737","journal-title":"Q Appl Math"},{"key":"4640_CR12","unstructured":"Gong C, Peng J, Liu Q (2019) Quantile stein variational gradient descent for batch Bayesian optimization. In: International Conference on machine learning, pp 2347\u20132356. PMLR."},{"key":"4640_CR13","unstructured":"Hein M, Bousquet O (2005) Hilbertian metrics and positive definite kernels on probability measures. In: International Workshop on Artificial Intelligence and Statistics, pp 136\u2013143. PMLR"},{"key":"4640_CR14","first-page":"20939","volume":"33","author":"N Jaquier","year":"2020","unstructured":"Jaquier N, Rozo L (2020) High-dimensional Bayesian optimization via nested Riemannian manifolds. Adv Neural Inf Process Syst 33:20939\u201320951","journal-title":"Adv Neural Inf Process Syst"},{"key":"4640_CR15","unstructured":"Jaquier N, Rozo L, Calinon S, B\u00fcrger M (2020) Bayesian optimization meets Riemannian manifolds in robot learning. In: Conference on Robot Learning, pp 233\u2013246. PMLR"},{"key":"4640_CR16","unstructured":"Kandasamy K, Schneider J, P\u00f3czos B (2015) High dimensional Bayesian optimisation and bandits via additive models. In: International Conference on machine learning, pp 295\u2013304. PMLR"},{"key":"4640_CR17","unstructured":"Kandasamy K, Neiswanger W, Schneider J, Poczos B, Xing EP (2018) Neural architecture search with bayesian optimisation and optimal transport.\u00a0In: Advances in neural information processing systems, p 31"},{"key":"4640_CR18","first-page":"199","volume":"37","author":"LV Kantorovich","year":"1942","unstructured":"Kantorovich LV (1942) On the translocation of masses. Dokl. Akad. Nauk. USSR (NS) 37:199\u2013201","journal-title":"Dokl. Akad. Nauk. USSR (NS)"},{"key":"4640_CR19","unstructured":"Lau TTK, Liu H (2022) Wasserstein distributionally robust optimization via Wasserstein barycenters. arXiv:2203.12136"},{"key":"4640_CR20","unstructured":"Liu Q, Wang D (2016) Stein variational gradient descent: a general purpose Bayesian inference algorithm. In: Advances in neural information processing systems, p 29"},{"key":"4640_CR35","unstructured":"Liu J, Wu J, Li B, Cui P (2022) Distributionally robust optimization with data geometry. In: Advances in neural information processing systems, vol 35, pp 33689\u201333701"},{"key":"4640_CR21","unstructured":"Liutkus A, Simsekli U, Majewski S, Durmus A, St\u00f6ter FR (2019) Sliced-Wasserstein flows: Nonparametric generative modeling via optimal transport and diffusions. In: International Conference on machine learning, pp 4104\u20134113. PMLR"},{"key":"4640_CR22","unstructured":"Monge G (1781) M\u00e9moire sur la th\u00e9orie des d\u00e9blais et des remblais. De l'Imprimerie Royale"},{"issue":"1","key":"4640_CR23","doi-asserted-by":"publisher","first-page":"51","DOI":"10.1007\/s11590-019-01433-w","volume":"14","author":"R Moriconi","year":"2020","unstructured":"Moriconi R, Kumar KS, Deisenroth MP (2020) High-dimensional Bayesian optimization with projections using quantile Gaussian processes. Optim Lett 14(1):51\u201364","journal-title":"Optim Lett"},{"key":"4640_CR24","doi-asserted-by":"crossref","unstructured":"Nguyen V, Gupta S, Rane S, Li C, & Venkatesh S (2017) Bayesian optimization in weakly specified search space. In: 2017 IEEE International Conference on data mining (ICDM), pp 347\u2013356. IEEE","DOI":"10.1109\/ICDM.2017.44"},{"issue":"3","key":"4640_CR25","doi-asserted-by":"publisher","DOI":"10.1088\/1751-8121\/ab5877","volume":"53","author":"K \u00d6cal","year":"2019","unstructured":"\u00d6cal K, Grima R, Sanguinetti G (2019) Parameter estimation for biochemical reaction networks using Wasserstein distances. J Phys A Math Theor 53(3):034002","journal-title":"J Phys A Math Theor"},{"issue":"5\u20136","key":"4640_CR26","doi-asserted-by":"publisher","first-page":"355","DOI":"10.1561\/2200000073","volume":"11","author":"G Peyr\u00e9","year":"2019","unstructured":"Peyr\u00e9 G, Cuturi M (2019) Computational optimal transport: with applications to data science. Found Trends\u00ae Mach Learn 11(5\u20136):355\u2013607","journal-title":"Found Trends\u00ae Mach Learn"},{"key":"4640_CR27","unstructured":"Redko I, Courty N, Flamary R, Tuia D (2019) Optimal transport for multi-source domain adaptation under target shift. In: The 22nd International Conference on artificial intelligence and statistics, pp 849\u2013858. PMLR"},{"key":"4640_CR28","unstructured":"Rout L, Korotin A, Burnaev E (2021) Generative modeling with optimal transport maps.\u00a0arXiv preprint arXiv:2110.02999"},{"key":"4640_CR29","first-page":"12356","volume":"33","author":"A Salim","year":"2020","unstructured":"Salim A, Korba A, Luise G (2020) The Wasserstein proximal gradient algorithm. Adv Neural Inf Process Syst 33:12356\u201312366","journal-title":"Adv Neural Inf Process Syst"},{"key":"4640_CR30","unstructured":"Vayer T, Chapel L, Flamary R, Tavenard R, Courty N (2018) Optimal transport for structured data with application on graphs.\u00a0arXiv preprint arXiv:1805.09114"},{"key":"4640_CR31","doi-asserted-by":"crossref","first-page":"23","DOI":"10.1007\/978-3-540-71050-9","volume-title":"Optimal transport: old and new","author":"C Villani","year":"2009","unstructured":"Villani C (2009) Optimal transport: old and new, vol 338. Springer, Berlin, p 23"},{"key":"4640_CR32","unstructured":"Zhang R, Chen C, Li C, Carin L (2018) Policy optimization as wasserstein gradient flows. In\u00a0International Conference on machine learning, pp 5737\u20135746. PMLR."},{"key":"4640_CR33","unstructured":"Zhang R, Wen Z, Chen C, Carin L (2019) Scalable Thompson sampling via optimal transport.\u00a0arXiv preprint arXiv:1902.07239."}],"container-title":["Journal of Ambient Intelligence and Humanized Computing"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s12652-023-04640-7.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1007\/s12652-023-04640-7\/fulltext.html","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s12652-023-04640-7.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,6,21]],"date-time":"2023-06-21T14:21:01Z","timestamp":1687357261000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1007\/s12652-023-04640-7"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,5,22]]},"references-count":35,"journal-issue":{"issue":"8","published-print":{"date-parts":[[2023,8]]}},"alternative-id":["4640"],"URL":"https:\/\/doi.org\/10.1007\/s12652-023-04640-7","relation":{},"ISSN":["1868-5137","1868-5145"],"issn-type":[{"value":"1868-5137","type":"print"},{"value":"1868-5145","type":"electronic"}],"subject":[],"published":{"date-parts":[[2023,5,22]]},"assertion":[{"value":"7 October 2022","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"8 May 2023","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"22 May 2023","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}}]}}