The Algorithm Selection Problem

By John R. Rice. 1976

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Optimal Decision Trees for the Algorithm Selection Problem: Integer Programming Based Approaches (2019, methodology): “From the observation that different algorithms may perform better on different classes of problems, Rice (1976) proposed a formal definition of the Algorithm Selection Problem (ASP).”
Fast Algorithm Selection Using Learning Curves (2015, methodology): “A common task, known as the Algorithm Selection Problem [13], is to determine which classifier performs best on a given dataset.”
Does Feature Selection Improve Classification? A Large Scale Experiment in OpenML (2016, background): “The algorithm selection problem, formalised by Rice in [17], is a natural problem from the field of meta-learning.”
Efficient Parameter Importance Analysis via Ablation with Surrogates (2017, methodology): “Before EPMs were used in algorithm configuration, they were already a key technique in perinstance algorithm selection (see, e.g., Rice 1976, Kotthoff 2014); furthermore, they have recently been used to define surrogate benchmarks for efficient benchmarking of hyperparameter optimization tools…”
Metalearning: a survey of trends and technologies (2013, background): “The idea of metalearning is not new, one of the first and seminal contributions having been provided by [53].”
Automatic performance tuning of sparse matrix kernels (2003, background): “The run-time selection problem for computational kernels was first posed by Rice [262], and again more recently by Brewer [56] (Section 9.”
Cross-disciplinary perspectives on meta-learning for algorithm selection (2009, background): “He concluded, however, that “the determination of the proper [nonlinear form is still somewhat of an art and there is no algorithm for making the choice” [Rice 1976].”