Decentralized Approximate Bayesian Inference for Distributed Sensor Network

Authors: Behnam Gholami, Sejong Yoon, Vladimir Pavlovic

AAAI 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental In this section, we first demonstrate the general convergence properties of the D-BPCA algorithm on synthetic data. We then apply our model to a set of Structure from Motion (Sf M) problems. We compared our distributed algorithm with traditional SVD, Centralized PCA (PPCA) (Tipping and Bishop 1999), Distributed PPCA (D-PPCA) (Yoon and Pavlovic 2012), and Centralized BPCA (BPCA) (Ilin and Raiko 2010).
Researcher Affiliation Academia Behnam Gholami, Sejong Yoon and Vladimir Pavlovic Rutgers, The State University of New Jersey 110 Frelinghuysen Road Piscataway, NJ 08854-8019 {bb510, sjyoon, vladimir}@cs.rutgers.edu
Pseudocode No The paper describes algorithms and derivations through mathematical equations and text, but it does not include any structured pseudocode or algorithm blocks.
Open Source Code No The paper does not provide any statement or link indicating that the source code for the methodology is openly available.
Open Datasets Yes We applied our model to the Caltech 3D Objects on Turntable dataset (Moreels and Perona 2007) and Hopkins155 dataset (Tron and Vidal 2007)
Dataset Splits No The paper mentions partitioning frames for distributed simulation and describes data usage for online learning, but it does not specify explicit training/validation/test dataset splits (e.g., percentages or sample counts) needed for reproduction.
Hardware Specification No The paper does not provide any specific hardware details such as GPU/CPU models, memory, or cloud computing specifications used for running its experiments.
Software Dependencies No The paper does not provide specific software dependencies with version numbers (e.g., library names with versions) that were used to conduct the experiments.
Experiment Setup Yes For all Sf M experiments, the network has the ring topology, with η = 10. We equally partitioned the frames into 5 nodes to simulate 5 cameras, the convergence was set to 10^-3 relative change in objective of (15).