.. glossary::
BO
Bayesian Optimization. A Black-Box optimization algorithm weighing exploration & exploitation
to find the minimum of its objective.
HB
`Hyperband <https://arxiv.org/abs/1603.06560>`_. A novel bandit-based algorithm for hyperparameter
optimization. Hyperband is an extension of successive halving and therefore works with
multi-fidelities.
BOHB
`Bayesian optimization and Hyperband <https://arxiv.org/abs/1807.01774>`_.
SMAC
Sequential Model-Based Algorithm Configuration.
ROAR
Random Online Adaptive Racing. A simple model-free instantiation of the general SMBO framework.
It selects configurations uniformly random and iteratively compares them against the current incumbent
using the intensification mechanism. See `SMAC extended <https://ai.dmi.unibas.ch/research/reading_group/hutter-et-al-tr2010.pdf>`_
chapter 3.2 for details.
BB
Black-Box. Refers to an algorithm being optimized, where only input and output are observable.
MF
Multi-Fidelity. Refers to running an algorithm on multiple budgets (such as number of epochs or
subsets of data) and thereby evaluating the performance prematurely.
TAE
Target Algorithm Evaluator. Your model, which returns a cost based on the given config,
budget and instance.
RF
Random Forest.
GP
Gaussian Process.
GP-MCMC
Gaussian Process with Markov-Chain Monte-Carlo.
Objective
An objective is a metric to evaluate the quality or performance of an Algorithm.
Budget
Budget is another word for fidelity. Examples are the number of training epochs or the size of
the data subset the algorithm is trained on.
PCS
`ConfigurationSpace <https://automl.github.io/ConfigSpace/master/API-Doc.html>`_ can be written/read from a PCS file.
EPM
Empirical Performance Models. Empirical performance models are regression models that characterize a given
algorithm’s performance across problem instances and/or parameter settings. These models can predict the
performance of algorithms on previously unseen input, including previously unseen problem instances and or
previously untested parameter settings and are useful for analyzing of how an algorithm performs under different
conditions, select promising configurations for a new problem instance, or surrogate benchmarks.
Intensification
A mechanism, that governs how many evaluations to perform with each configuration and when to trust a configuration
enough to make it the new current best known configuration (the incumbent).
CV
Cross-Validation.
CLI
Command-Line Interface.