A character vector of arguments (character strings) of the form "<name>=<value>".
Values will be converted to logical or numeric when possible.
Accepted <names> are below. Defaults in parenthesis:
- decomp
Either 'CP' or 'Tucker'. (Tucker)
- row.share
Logical. Should the variance be shared across rows of the projection matrices?
This will cause predictors to be or excluded for the whole model, instead of just for particular
latent factors. (F)
- seed
Numeric. Seed used for random initialization. (NA)
- verbose
Logical. Display more messages during training. (F)
- parallel
Logical. Perform operations in parallel when possible. (T)
- cores
Numeric. The number of parallel threads to use. (2)
- lower.bnd
Logical. Should the lower bound be calculated during training. Setting to FALSE
saves time (F)
- RMSE
Logical. Should the root mean squared error for the training data be calculated during
training. (T)
- cor
Logical. Should the Pearson correlation for the training data be calculated during
training. (T)
- A1.intercept
Logical. Add a constant column to the mode 1 predictors. (T)
- A2.intercept
Logical. Add a constant column to the mode 2 predictors. (T)
- A3.intercept
Logical. Add a constant column to the mode 3 predictors. (F)
- H1.intercept
Logical. Add a constant column to the mode 1 latent (H) matrix. (F)
- H2.intercept
Logical. Add a constant column to the mode 2 latent (H) matrix. (F)
- H3.intercept
Logical. Add a constant column to the mode 3 latent (H) matrix. (F)
- R
Numeric. Number of latent factors used in a CP model. (3)
- R1
Numeric. Number of latent factors used for mode 1 in a Tucker decomposition. (3)
- R2
Numeric. Number of latent factors used for mode 2 in a Tucker decomposition. (3)
- R3
Numeric. Number of latent factors used for mode 3 in a Tucker decomposition. (3)
- core.updates
Numeric. Number of core elements to update each round for stochastic training. (all)
- m1.alpha
Numeric. Prior for the 'shape' parameter of the gamma distribution on the
precision values in the mode 1 projection (A) matrix. Set this to a small value (ex. 1e-10)
to encourage sparsity in mode 1 predictors. (1)
- m2.alpha
Numeric. Same as above for mode 2. (1)
- m3.alpha
Numeric. Same as above for mode 3. (1)
- m1.beta
Numeric. Prior for the 'scale' parameter of the gamma distribution on the
precision values in the mode 1 projection (A) matrix. Set this to a large value (ex. 1e10)
to encourage sparsity in mode 1 predictors. Note this should stay balanced with m1.alpha
so thir product is 1. (1)
- m2.beta
Numeric. Same as above for mode 2. (1)
- m3.beta
Numeric. Same as above for mode 3. (1)
- A.samp.sd
Numeric. Standard deviation used when initializing values in the projection
(A) matrices. (1)
- H.samp.sd
Numeric. Standard deviation used when initializing values in the latent
(H) matrices. (1)
- R.samp.sd
Numeric. Standard deviation used when initializing values in the core
tensor for Tucker models. (1)
- A.var
Numeric. Initial variance for projection (A) matrices. (1)
- H.var
Numeric. Initial variance for latent (H) matrices. (1)
- R.var
Numeric. Initial variance for the core tensor in Tucker models. (1)
- random.H
Logical. Should the latent matrices be initialized randomly or be the result
of multiplying the input data by the projection matrices. (T)
- core.0D.alpha
Numeric. Prior for the 'scale' parameter of the gamma distribution on the
precision value in the element of the core tensor corresponding to the intercept for all
three modes (core.mean[1,1,1]). Only used for Tucker models when all H intercepts are true.
Set this to a small value (ex. 1e-10) to encourage sparsity in core predictor. (1)
- core.1D.alpha
Numeric. As above for values corresponding to the intercepts for
two modes (core.mean[1,1,], core.mean[1,,1] and core.mean[,1,1]). (1)
- core.2D.alpha
Numeric. As above for values corresponding to the intercepts for
one mode (core.mean[1,,], core.mean[,1,] and core.mean[,,1]). (1)
- core.3D.alpha
Numeric. As above for values not corresponding to intercepts. (1)
- core.0D.beta
Numeric. As above but a prior for the 'scale' parameter. (1)
- core.1D.beta
Numeric. As above but a prior for the 'scale' parameter. (1)
- core.2D.beta
Numeric. As above but a prior for the 'scale' parameter. (1)
- core.3D.beta
Numeric. As above but a prior for the 'scale' parameter. (1)
- m1.sigma2
Numeric. Variance for the mode 1 latent (H) matrix. Set small to link the
values in the latent matrices to the product of the input and projection matrices. If there
is no input data, set to one or larger. (0.01)
- m2.sigma2
Numeric. As above for mode 2. (0.01)
- m3.sigma2
Numeric. As above for mode 3. (0.01)
- sigma2
Numeric. Variance parameter for the response tensor or 'auto' (default).
If set to 'auto' then the square-root of the variance of the training responses is used.
- remove.start
Numeric. The iteration to begin removing predictors if any of
m1.remove.lmt
, m2.remove.lmt
, m3.remove.lmt
or remove.per
are set. (Inf)
- remove.per
Numeric. Percentage of predictors to remove with the lowest mean of
squared values across rows of the projection matrix. (0)
- m1.remove.lmt
Numeric. Remove a mode 1 predictor if the mean squared value of
its row in the projection matrix drop below this value. (0)
- m2.remove.lmt
As above for mode 2. (0)
- m3.remove.lmt
As above for mode 3. (0)
- early.stop
Numeric. Stop training if the lower bound value changes by less than
this value. (0)
- plot
Logical. Show plots while training
- show.mode
Numeric vector. Display images of the projection and latent matrices
for these modes while training. (c(1,2,3))
- update.order
Numeric vector. Update the modes in this order (c(3,2,1))