Cascaded Algorithm-Selection and Hyper-Parameter Optimization with Extreme-Region Upper Confidence Bound Bandit

Authors: Yi-Qi Hu, Yang Yu, Jun-Da Liao

IJCAI 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We also conduct experiments on a synthetic problem as well as a set of Auto ML tasks. The results verify the effectiveness of the proposed method. The experiments on synthetic and real Auto ML tasks reveal that the ER-UCB can find the best algorithm precisely, and exploit it with the majority of the trial budget.
Researcher Affiliation Academia Yi-Qi Hu , Yang Yu and Jun-Da Liao National Key Laboratory for Novel Software Technology, Nanjing University, Nanjing 210023, China {huyq, yuy}@lamda.nju.edu.cn, liaojd98@gmail.com
Pseudocode Yes Algorithm 1 Extreme-region UCB Bandit
Open Source Code No No explicit statement or link regarding open-source code availability for the methodology described in this paper was found.
Open Datasets Yes And 12 classification datasets from UCI are selected as Auto ML tasks.
Dataset Splits No Let Dtrain and Dtest denote the training and testing datasets... The Auto ML problem can be formulated as follows: ... where Dvalid j Dtrain and Dtrain j = Dtrain Dvalid j. (While validation data is mentioned, specific split percentages or methodology for creating the splits from the datasets are not provided.)
Hardware Specification No No explicit hardware specifications (e.g., specific CPU/GPU models or memory) used for running experiments were provided in the paper.
Software Dependencies No We select 10 frequently-used algorithms as the candidates from SKLEARN [Pedregosa et al., 2011] (Mentions SKLEARN but no version number.)
Experiment Setup Yes The trial budget is set as 1000. The experiment for every hyperparameter setting is repeated for 3 times independently, and the average results are presented. The trial budget is 1000. We set θ = 0.01, γ = 20 for the ER-UCB on all datasets. The β is set according to the tasks, and showed in Table 2. For each method and each dataset, we run every experiment 3 times independently, and the average performances of our experiment are presented.