Collaborative Learning for Deep Neural Networks

Authors: Guocong Song, Wei Chai

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

Reproducibility Variable Result LLM Response
Research Type Experimental The empirical results on CIFAR and Image Net datasets demonstrate that deep neural networks learned as a group in a collaborative way significantly reduce the generalization error and increase the robustness to label noise.
Researcher Affiliation Industry Guocong Song Playground Global Palo Alto, CA 94306 songgc@gmail.com Wei Chai Google Mountain View, CA 94043 chaiwei@google.com
Pseudocode No The paper describes procedures and optimizations in paragraph form but does not include any structured pseudocode or algorithm blocks.
Open Source Code No The paper does not contain any statements about releasing code for the described methodology or provide a link to a code repository.
Open Datasets Yes The two CIFAR datasets, CIFAR-10 and CIFAR-100, consist of colored natural images with 32x32 pixels [13]... The ILSVRC 2012 classification dataset consists of 1.2 million for training, and 50,000 for validation [6].
Dataset Splits Yes The ILSVRC 2012 classification dataset consists of 1.2 million for training, and 50,000 for validation [6]. ... Table 4: Validation errors of Res Net-50 on Image Net.
Hardware Specification No The paper mentions 'graphics processing unit (GPU)' generally and discusses 'GPU memory consumption' but does not specify any particular GPU model (e.g., NVIDIA, A100) or other hardware components like CPU type or memory size.
Software Dependencies No The paper states 'All experiments are conducted with Tensorflow [1].' but does not provide a specific version number for TensorFlow or any other software dependencies used.
Experiment Setup Yes We use T = 2 and β = 0.5 for all experiments. ... Refer to Section 2 in Supplementary material for the detailed training setup.