shin92train: Input representation of shin92 for models input-compatible...

In catlearn: Formal Psychological Models of Categorization and Learning

Input representation of shin92 for models input-compatible with slpALCOVE.

Description

Creates randomized training and transfer blocks for CIRP shin92 , in a format suitable for the slpALCOVE model, and any other model that uses the same input representation format. The stimulus co-ordinates come from a MDS solution reported by Shin & Nosofsky (1992).

Usage

  shin92train(condition = 'equal3', learn.blocks = 8, trans.blocks = 3,
          absval = -1, format = 'mds', subjs = 1, seed = 8416, missing =
          'geo')

Arguments

condition	Experimental condition 'equal3', 'equal10', 'unequal3', or 'unequal10', as defined by Shin & Nosofsky (1992).
learn.blocks	Number of training blocks to generate. Omit this argument to get the same number of training blocks as the published study (8).
trans.blocks	Number of transfer blocks to generate. Omit this argument to get the same number of transfer blocks as the published study (3).
absval	Teaching value to be used where category is absent.
format	Specifies format used for input representation. Only one format is currently suported, so this option is provided solely to support future development.
subjs	Number of simulated subjects to be run.
seed	Sets the random seed
missing	If set to 'geo', output missing dimension flags (see below)

Details

A matrix is produced, with one row for each trial, and with the following columns:

ctrl - Set to 1 (reset model) for trial 1, set to zero (normal trial) for all other training trials, and set to 2 (freeze learning) for all transfer trials.

cond - 1 = equal3, 2 = equal10, 3 = unequal3, 4 = unequal10

phase - 1 = training, 2 = transfer

blk - block of trials

stim - stimulus number; these correspond to the rows in Tables A3 and A4 of Shin & Nosofsky (1992)

x1 ... x6 - input representation. These are the co-ordinates of an MDS solution for these stimuli (see Shin & Nosofsky, 1992, Tables A3 and A4). Note: Size 3 conditions have a four-dimensional MDS solution, so the output is x1 ... x4

t1, t2 - teaching signal (1 = category present, absval = category absent)

m1 ... m6 - Missing dimension flags (always set to zero in this experiment, indicating all input dimensions are present on all trials). Note: ranges from m1 to m4 for Size 3 conditions. Only produced if missing = 'geo'.

Although the trial ordering is random, a random seed is used, so multiple calls of this function with the same parameters should produce the same output. This is usually desirable for reproducibility and stability of non-linear optimization. To get a different order, use the seed argument to set a different seed.

This function was originally developed to support simulations reported in Wills et al. (2017).

Value

R by C matrix, where each row is one trial, and the columns contain model input.

Author(s)

Andy Wills

References

Shin, H.J. & Nosofsky, R.M. (1992). Similarity-scaling studies of dot-pattern classification and recognition. Journal of Experimental Psychology: General, 121, 278-304.

Wills, A.J., O'Connell, G., Edmunds, C.E.R. & Inkster, A.B. (2017). Progress in modeling through distributed collaboration: Concepts, tools, and category-learning examples. The Psychology of Learning and Motivation, 66.

See Also

shin92, shin92oat, slpALCOVE

catlearn documentation built on April 4, 2023, 5:12 p.m.