The Label Complexity of Active Learning from Observational Data

Authors: Songbai Yan, Kamalika Chaudhuri, Tara Javidi

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

Reproducibility Variable Result LLM Response
Research Type Experimental We present some empirical results in Section F in Appendix.
Researcher Affiliation Academia Songbai Yan University of California San Diego yansongbai@eng.ucsd.edu Kamalika Chaudhuri University of California San Diego kamalika@cs.ucsd.edu Tara Javidi University of California San Diego tjavidi@eng.ucsd.edu
Pseudocode Yes Algorithm 1 Disagreement-Based Active Learning with Logged Observational Data
Open Source Code No The paper does not provide an explicit statement about releasing its source code nor does it include a link to a repository for the code described in the paper.
Open Datasets No The paper defines abstract datasets like 'T0 = {(Xt, Yt, Zt)}m t=1' and 'additional n examples {(Xt, Yt)}m+n t=m+1 drawn i.i.d. from distribution D,' but it does not specify a concrete, publicly available dataset with access information (link, citation, or repository).
Dataset Splits No The paper discusses data collection ('logged observational data' and 'online stream') but does not provide specific details on training, validation, or test set splits, percentages, or sample counts within the main text.
Hardware Specification No The paper does not provide any specific hardware details such as GPU models, CPU types, or cloud computing instance specifications used for running experiments.
Software Dependencies No The paper references 'Vowpal Wabbit' but does not specify version numbers for any software components or libraries used in its methodology or experiments.
Experiment Setup Yes Input: confidence δ, logged data T0, epoch schedule τ1, . . . , τK, n = PK i=1 τi.