Gaussian Processes for Survival Analysis

Authors: Tamara Fernandez, Nicolas Rivera, Yee Whye Teh

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

Reproducibility Variable Result LLM Response
Research Type Experimental We report experimental results on synthetic and real data, showing that our model performs better than competing models such as Cox proportional hazards, ANOVA-DDP and random survival forests.
Researcher Affiliation Academia Tamara Fernández Department of Statistics, University of Oxford. Oxford, UK. fernandez@stats.ox.ac.uk Nicolas Rivera Department of Informatics, King s College London. London, UK. nicolas.rivera@kcl.ac.uk Yee Whye Teh Department of Statistics, University of Oxford. Oxford, UK. y.w.teh@stats.ox.ac.uk
Pseudocode Yes Algorithm 1: Inference Algorithm.
Open Source Code No The paper does not provide any explicit statements about releasing source code for the described methodology, nor does it include a link to a code repository.
Open Datasets Yes We run experiments on the Veteran data, avaiable in the R-package survival package [19].
Dataset Splits Yes We perform 10-fold cross validation and compute the C-index for each fold.
Hardware Specification No The paper does not provide any specific details about the hardware (e.g., CPU, GPU models, memory) used to run the experiments.
Software Dependencies No The paper mentions using the 'R-package survival package [19]' but does not specify its version number or any other software dependencies with version information.
Experiment Setup Yes All the experiments are performed using our approximation scheme of equation (6) with a value of m = 50. ... We choose a gamma prior on β and a uniform U(0, 2.3) on α. ... We use elliptical slice sampler [16] for jointly updating the set of coefficients {aj k, bj k} and length-scale parameters. ... We use the maximum likelihood estimator as initial parameters of the model.