Active Heteroscedastic Regression

Authors: Kamalika Chaudhuri, Prateek Jain, Nagarajan Natarajan

ICML 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We conclude the paper presenting simulations supporting our theoretical bounds as well as experiments on real-world data.
Researcher Affiliation Collaboration Kamalika Chaudhuri 1 Prateek Jain 2 Nagarajan Natarajan 2 ... 1University of California, San Diego 2Microsoft Research, India.
Pseudocode Yes Algorithm 1 Passive Regression With Noise Oracle; Algorithm 2 Active Regression With Noise Oracle; Algorithm 3 Least Squares with Estimated Weights; Algorithm 4 Active Regression
Open Source Code No The paper does not provide any concrete access to source code for the methodology described.
Open Datasets Yes We also study the performance of the algorithms on two real-world datasets from UCI: (1) WINE QUALITY with m = 6500 and d = 11, and (2) MSD (a subset of the million song dataset) with m = 515345 and d = 90.
Dataset Splits No For each dataset, we create a 70-30 train-test split, and learn the best linear regressor using ordinary least squares, which forms our β .
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 does not provide specific ancillary software details with version numbers (e.g., library or solver names with version numbers).
Experiment Setup No The paper describes the synthetic data generation process and data splitting for real-world datasets, but it does not specify concrete hyperparameter values or detailed training configurations for the models (e.g., learning rates, batch sizes, optimizers).