Using Anytime Algorithms in Intelligent Systems
By Shlomo Zilberstein. 1996
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
- Learning Algorithms for Active Learning (2017, methodology): “In this sense, we build an anytime active learner (Zilberstein, 1996), with the model trained at each step to output the best possible prediction on the evaluation set.”
- Automatically improving the anytime behaviour of optimisation algorithms (2014, methodology): “Definition 1 (Performance profile [58]).”
- Anytime Ranking for Impact-Ordered Indexes (2015, background): “Anytime algorithms are algorithms where the quality of results improves as the computation time increases [21].”
- Anytime Heuristic Search (2011, background): “…under varying or uncertain time constraints because they have a solution ready whenever they are stopped (with the possible exception of an initial time period before the first solution is found) and the quality of the solution improves with additional computation time (Zilberstein, 1996).”
- Book Review: Computational Methods of Feature Selection (2007, background): “Algorithms that generate initial solutions quickly, and then improve them over time are also known as anytime algorithms [22].”
- Neo: A Learned Query Optimizer (2019, background): “d explored. While one could terminate this search process as soon as a leaf node (a complete execution plan) is found, this search procedure can easily be transformed into an anytime search algorithm [64], i.e. an algorithm that continues to nd better and better results until a xed time cuto. In thi”
- Distillation-Based Training for Multi-Exit Architectures (2019, background): “In [6], anytime algorithms are defined for the first time, and they become widely popular in planning and control [8, 29, 49].”
- SpeedBoost: Anytime Prediction with Uniform Near-Optimality (2012, background): “Anytime algorithms [18] exhibit this desirable trait of providing increasingly better results given more computation time.”
- A hybrid genetic algorithm for feature selection wrapper based on mutual information (2007, background): “…of avoiding getting stuck in local optima often encountered in feature selection problems; second, they may be classified into a kind of anytime algorithms (Zilberstein, 1996), which can generate currently best sub-sets constantly and keep improving the quality of selected features as time goes on.”
- Fairness-aware Learning through Regularization Approach (2011, background): “Regarding underestimation, the concepts of anytime algorithms in planning or decision making [18] might be useful.”