Research map / Probabilistic and scientific modeling
Bayesian optimization: research map
434 papers on Bayesian optimization in Probabilistic and scientific modeling, from ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR (2016–2026), grouped into 2 clusters and 7 approaches. The busiest year so far is 2026.
What it is
Some functions are expensive to evaluate: a lab experiment, a long training run. Bayesian optimization finds a good input in few tries by keeping a probabilistic model of the function and choosing the next point that best balances exploring and exploiting; hyperparameter tuning is a common use.
Terms you will see on this page:
- Surrogate model
- a cheap model (often a Gaussian process) of the expensive function.
- Acquisition function
- the rule that picks the next point to evaluate.
- Black-box optimization
- optimising a function you can only query, without gradients.
Start reading with the most cited papers here:
- BatchBALD: Efficient and Diverse Batch Acquisition for Deep Bayesian Active Learning (NeurIPS 2019 · 294 citations)
- Bayesian Optimization with Robust Bayesian Neural Networks (NeurIPS 2016 · 263 citations)
Within Probabilistic and scientific modeling, its share held steady from 4.5% in 2023–24 to 4.4% in 2025–26 (92 → 134 papers at ICML, NeurIPS, ICLR, CVPR and AAAI, the venues with data for all four years).
Explore Bayesian optimization in the interactive map
Working on something in this topic? Describe your idea in scime atlas to see which approach it falls under, the closest papers by meaning and how crowded the spot has become.
What changed recently
Approaches whose share of the topic moved most between 2023–24 and 2025–26 (ICML, NeurIPS, ICLR, CVPR and AAAI):
- Offline black-box optimization with surrogates rising: 11% → 22% of the topic (10 → 30 papers)
- Local and high-dimensional Bayesian optimization rising: 16% → 28% of the topic (15 → 37 papers)
- Multi-objective Bayesian optimization rising: 8% → 10% of the topic (7 → 14 papers)
- Regret bounds for Bayesian optimization falling: 16% → 8% of the topic (15 → 11 papers)
- Acquisition function design falling: 28% → 22% of the topic (26 → 29 papers)
Papers from 2025–2026 that are already cited most:
- ROOT: Rethinking Offline Optimization as Distributional Translation via Probabilistic Bridge (NeurIPS 2025 · 1 citations)
- Context in Public Health for Underserved Communities: A Bayesian Approach to Online Restless Bandits (AAAI 2025 · 1 citations)
Approaches and key papers
Bayesian optimization methods · 338 papers
Keywords: acquisition, bayesian, multi objective
Approaches in this cluster:
- Acquisition function design (106 papers)
Develop new acquisition criteria and stopping rules for sequential Bayesian optimization. - Local and high-dimensional Bayesian optimization (96 papers)
Make Bayesian optimization scale with local search, structure, and analyses of high-dimensional behavior. - Multi-fidelity hyperparameter optimization (56 papers)
Tune hyperparameters efficiently by allocating budget across fidelities and early stopping. - Multi-objective Bayesian optimization (35 papers)
Approximate Pareto fronts with batch, diversity, and preference-aware acquisition. - Regret bounds for Bayesian optimization (45 papers)
Prove regret guarantees for Bayesian and preferential optimization in kernel bandits.
Most cited and most cited since 2024:
- BatchBALD: Efficient and Diverse Batch Acquisition for Deep Bayesian Active Learning (NeurIPS 2019 · 294 citations)
- Bayesian Optimization with Robust Bayesian Neural Networks (NeurIPS 2016 · 263 citations)
- Constrained Bayesian Optimization under Partial Observations: Balanced Improvements and Provable Convergence (AAAI 2024 · 15 citations)
- Are You Concerned about Limited Function Evaluations: Data-Augmented Pareto Set Learning for Expensive Multi-Objective Optimization (AAAI 2024 · 11 citations)
Black-box and offline optimization · 96 papers
Keywords: black box, box optimization, offline
Approaches in this cluster:
- Offline black-box optimization with surrogates (62 papers)
Optimize designs from static datasets using learned surrogates, generative models, and language-model embeddings. - High-dimensional Bayesian optimization (34 papers)
Scale Bayesian optimization to high-dimensional and discrete spaces using structured kernels and search-space learning.
Most cited and most cited since 2024:
- Multi-fidelity Bayesian Optimisation with Continuous Approximations (ICML 2017 · 104 citations)
- Scalable Hyperparameter Transfer Learning (NeurIPS 2018 · 77 citations)
- A survey and benchmark of high-dimensional Bayesian optimization of discrete sequences (NeurIPS 2024 · 3 citations)
- Grey-Box Bayesian Optimization for Sensor Placement in Assisted Living Environments (AAAI 2024 · 3 citations)
Related topics in Probabilistic and scientific modeling
- Dynamical systems (1,019)
- Uncertainty and Gaussian processes (1,421)
- Molecules and proteins (2,120)
- Variational and generative sampling (2,370)
- Treatment effects (468)
- Differential equations and PDEs (676)
- Causal discovery (973)
