micp.stats: Mixed-effects inference on classification performance

In kaybrodersen/micp: Bayesian mixed-effects inference on classification performance

Mixed-effects inference on classification performance

Description

Mixed-effects inference on classification performance

Usage

micp.stats(ks, ns)

Arguments

ks	When inferring on accuracies: m-sized vector of number of correct predictions (in each subject). When inferring on balanced accuracies: 2xm matrix of correctly predicted positive trials (first row) and corectly predicted negative trials (second row).
ns	When inferring on accuracies: m-sized vector of total number of trials (in each subject). When inferring on balanced accuracies: 2xm matrix of total number of positive (first row) and negative (second row) trials.

Value

A list with the following fields: mu, p, ci, and stats.

mu: Posterior mean of the population mean accuracy or balanced accuracy. This is the expected performance of the classifier at the group level.

p: Posterior infraliminal probability of the population mean. This is the posterior belief that the classifier did not operate above chance (50 above-chance performance.

ci: Posterior 95 can be used to show error bars around mu.

stats: Additional return values, depending on the selected model. See individual inference functions for details.

Author(s)

Kay H. Brodersen, ETH Zurich

References

K.H. Brodersen, J. Daunizeau, C. Mathys, J.R. Chumbley, J.M. Buhmann, & K.E. Stephan (2013). Variational Bayesian mixed-effects inference for classification studies. NeuroImage (in press). doi:10.1016/j.neuroimage.2013.03.008.

K.H. Brodersen, C. Mathys, J.R. Chumbley, J. Daunizeau, C.S. Ong, J.M. Buhmann, & K.E. Stephan (2012). Bayesian mixed-effects inference on classification performance in hierarchical datsets. Journal of Machine Learning Research, 13, 3133-3176.

K.H. Brodersen, C.S. Ong, J.M. Buhmann, & K.E. Stephan (2010). The balanced accuracy and its posterior distribution. ICPR, 3121-3124.

Examples

  # Accuracy:
  ks <- c(19, 41, 15, 39, 39)
  ns <- c(45, 51, 20, 46, 58)
  results1 <- micp::micp.stats(ks, ns)
  print(results1)

  # Balanced accuracy:
  ks <- rbind(c(19, 41, 15, 39, 39), c(41, 46, 43, 48, 37))
  ns <- rbind(c(45, 51, 20, 46, 58), c(55, 49, 80, 54, 42))
  results2 <- micp::micp.stats(ks, ns)
  print(results2)

kaybrodersen/micp documentation built on April 15, 2022, 2:24 a.m.