Information Gathering With Peers: Submodular Optimization With Peer-Prediction Constraints

Authors: Goran Radanovic, Adish Singla, Andreas Krause, Boi Faltings

AAAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Finally, we evaluate our methods using a realistic crowd sensing testbed. ... We perform two different tests. In the first test, we vary the total available budget B from 5 to 25 at steps of size 5. At the same time, we keep the minimal expected payment to τmin = 0.5. As we can see from Figure 2a, as the total budget increases, all the methods perform better. ... The results are given in Figure 2b.
Researcher Affiliation Academia Goran Radanovic Harvard University Cambridge, USA gradanovic@g.harvard.edu Adish Singla MPI-SWS Saarbr ucken, Germany adishs@mpi-sws.org Andreas Krause ETH Zurich Zurich, Switzerland krausea@ethz.ch Boi Faltings EPFL Lausanne, Switzerland boi.faltings@epfl.ch
Pseudocode Yes Algorithm 1: Algorithm PPCGREEDY... Algorithm 2: Algorithm PPCGREEDYITER
Open Source Code No The paper does not include any explicit statements or links indicating the release of open-source code for the described methodology.
Open Datasets Yes We use a crowd sensing test-bed of Singla (2017), constructed from real measurements of CO2 and user locations across an urban area. ... obtained using a publicly available data (Open Street Map5), from which we randomly selected 300 locations from an area in the center of New York City. ... inferred from a publicly accessible dataset (Strava6) that contains the mobility patterns of cyclists for a period of 6 days.
Dataset Splits No The paper describes how the dataset was collected and sampled but does not specify a training, validation, or test split for the data used in experiments.
Hardware Specification No The paper does not provide specific details about the hardware used to run the experiments, such as CPU/GPU models or cloud instance types.
Software Dependencies No The paper does not specify any software dependencies with version numbers, such as programming languages or libraries.
Experiment Setup Yes In the first test, we vary the total available budget B from 5 to 25 at steps of size 5. At the same time, we keep the minimal expected payment to τmin = 0.5. ... In the second test, we vary the minimal expected payment τmin, which now takes values in {0.1, 0.25, 0.5, 0.75, 0.9}. Budget B is set to 15.