Shift-Pessimistic Active Learning Using Robust Bias-Aware Prediction

Authors: Anqi Liu, Lev Reyzin, Brian Ziebart

AAAI 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We investigate the theoretical benefits of this approach and demonstrate its empirical advantages on probabilistic binary classification tasks. In addition to these theoretical properties, we evaluate and compare the effectiveness of our approach on a range of classification tasks. Experiments Classification tasks We evaluate the performance of different active learning approaches using four datasets from the UCI repository (Bache and Lichman 2013).
Researcher Affiliation Academia Anqi Liu Department of Computer Science University of Illinois at Chicago Chicago, IL 60607 aliu33@uic.edu; Lev Reyzin Department of Mathematics, Statistics, and Computer Science University of Illinois at Chicago Chicago, IL 60607 lreyzin@math.uic.edu; Brian D. Ziebart Department of Computer Science University of Illinois at Chicago Chicago, IL 60607 bziebart@uic.edu
Pseudocode Yes Algorithm 1 Label solicitation for pool-based active learner with covariate shift correction
Open Source Code No The paper does not provide an explicit statement or link for open-source code for the described methodology.
Open Datasets Yes We evaluate the performance of different active learning approaches using four datasets from the UCI repository (Bache and Lichman 2013).
Dataset Splits No The paper specifies a training and testing split ('80% of data' for training, 'remaining 20%' for testing), but does not explicitly mention a separate validation split with percentages or sample counts.
Hardware Specification No The paper does not provide specific hardware details (e.g., GPU/CPU models, memory) used for running its experiments.
Software Dependencies No The paper mentions techniques like 'logistic regression' and 'Gaussian kernel density estimation' but does not list specific software packages or libraries with version numbers.
Experiment Setup Yes For all methods, we use first-order and second-order statistics of the inputs as features... we use a different regularization weight for each feature corresponding with the 95% confidence interval of the feature s mean... We apply Gaussian kernel density estimation (KDE) on the labeled examples to estimate the labeled data density... For higher dimensional data (Seed and E. coli), we first apply principal component analysis to reduce the dimensionality to a space that covers at least 95% of the input variance.