The function oracle performs a strategie that cannot be defined online
(in contrast to mixture). It requires in advance the knowledge of the whole
data set Y and the expert advice to be well defined.
Examples of oracles are the best fixed expert, the best fixed convex
combination rule, the best linear combination rule, or the best expert
that can shift a few times.
oracle(
Y,
experts,
model = "convex",
loss.type = "square",
awake = NULL,
lambda = NULL,
niter = NULL,
...
)# S3 method for oracle
plot(x, sort = TRUE, col = NULL, ...)
A vector containing the observations to be predicted.
A matrix containing the experts
forecasts. Each column corresponds to the predictions proposed by an expert
to predict Y. It has as many columns as there are experts.
A character string specifying the oracle to use or a list with a component name specifying the oracle and any additional parameter needed.
Currently available oracles are:
The best fixed (constant over time) expert oracle.
The best fixed convex combination (vector of non-negative weights that sum to 1)
The best fixed linear combination of expert
It computes for all number $m$ of stwitches the
sequence of experts with at most $m$ shifts that would have performed the
best to predict the sequence of observations in Y.
A string or a list with a component 'name' specifying the loss function considered to evaluate the performance. It can be 'square', 'absolute', 'percentage', or 'pinball'. In the case of the pinball loss, the quantile can be provided by assigning to loss.type a list of two elements:
A string defining the name of the loss function (i.e., 'pinball')
A number in [0,1] defining the quantile to be predicted. The default value is 0.5 to predict the median.
A matrix specifying the
activation coefficients of the experts. Its entries lie in [0,1].
Possible if some experts are specialists and do not always form and suggest
prediction. If the expert number k at instance t does not
form any prediction of observation Y_t, we can put
awake[t,k]=0 so that the mixture does not consider expert k in
the mixture to predict Y_t. Remark that to compute the best expert oracle,
the performance of unactive (or partially active) experts is computed by using
the prediction of the uniform average of active experts.
A positive number used by the 'linear' oracle only. A possible $L_2$ regularization parameter for computing the linear oracle (if the design matrix is not identifiable)
A positive integer for 'convex' and 'linear' oracles if direct computation of the oracle is not implemented. It defines the number of optimization steps to perform in order to approximate the oracle (default value is 3).
Additional parameters
that are passed to optim function is order to perform convex optimization
(see parameter niter).
An object of class oracle.
if set to TRUE (default), it sorts the experts by performance before the plots.
colors
An object of class 'oracle' that contains:
The average loss suffered by the oracle. For the 'shifting' oracle,
it is a vector of length T where
T is the number of instance to be predicted (i.e., the length of the
sequence Y). The value of $loss(m)$ is the loss
(determined by the parameter loss.type) suffered by the
best sequence of expert with at
most $m-1$ shifts.
Not for the 'shifting' oracle. A vector containing the best weight vector corresponding to the oracle.
Not for the 'shifting' oracle. A vector containing the predictions of the oracle.
If loss.type is the square loss (default) only.
The root mean square error (i.e., it is the square root of loss.
oracle: plot. It has one optional arguments.