wconf (version 1.2.0)

wconfusionmatrix: Weighted confusion matrix

Description

This function calculates the weighted confusion matrix from a caret ConfusionMatrix object or a simple matrix, according to one of several weighting schemas and optionally prints the weighted accuracy score.

Usage

wconfusionmatrix(m, weight.type = "arithmetic",
                        weight.penalty = FALSE,
                        standard.deviation = 2,
                        geometric.multiplier = 2,
                        interval.high=1, interval.low = -1,
                        sin.high=1.5*pi, sin.low = 0.5*pi,
                        tanh.decay = 3,
                        custom.weights = NA,
                        print.weighted.accuracy = FALSE)

Value

an nxn weighted confusion matrix

Arguments

m

the caret confusion matrix object or simple matrix.

weight.type

the weighting schema to be used. Can be one of: "arithmetic" - a decreasing arithmetic progression weighting scheme, "geometric" - a decreasing geometric progression weighting scheme, "normal" - weights drawn from the right tail of a normal distribution, "interval" - weights contained on a user-defined interval, "sin" - a weighing scheme based on a sine function, "tanh" - a weighing scheme based on a hyperbolic tangent function, "custom" - custom weight vector defined by the user.

weight.penalty

determines whether the weights associated with non-diagonal elements generated by the "normal", "arithmetic" and "geometric" weight types are positive or negative values. By default, the value is set to FALSE, which means that generated weights will be positive values.

standard.deviation

standard deviation of the normal distribution, if the normal distribution weighting schema is used.

geometric.multiplier

the multiplier used to construct the geometric progression series, if the geometric progression weighting scheme is used.

interval.high

the upper bound of the weight interval, if the interval weighting scheme is used.

interval.low

the lower bound of the weight interval, if the interval weighting scheme is used.

sin.high

the upper segment of the sine function to be used in the weighting scheme.

sin.low

the lower segment of the sine function to be used in the weighting scheme.

tanh.decay

the decay factor of the hyperbolic tangent weighing function. Higher values increase the rate of decay and place less weight on observations farther away from the correctly predicted category.

custom.weights

the vector of custom weight to be applied, if the custom weighting scheme was selected. The vector should be equal to "n", but can be larger, with excess values being ignored.

print.weighted.accuracy

print the weighted accuracy metric, which represents the sum of all weighted confusion matrix cells divided by the total number of observations.

Author

Alexandru Monahov, <https://www.alexandrumonahov.eu.org/>

Details

The number of categories "n" should be greater or equal to 2.

See Also

[weightmatrix()] for the weight matrix used in computations, [balancedaccuracy()] for accuracy metrics designed for imbalanced data.

Examples

m = matrix(c(70,0,0,10,10,0,5,3,2), ncol = 3, nrow=3)
wconfusionmatrix(m, weight.type="arithmetic", print.weighted.accuracy = TRUE)
wconfusionmatrix(m, weight.type="geometric", print.weighted.accuracy = TRUE)
wconfusionmatrix(m, weight.type="interval", print.weighted.accuracy = TRUE)
wconfusionmatrix(m, weight.type="normal", print.weighted.accuracy = TRUE)
wconfusionmatrix(m, weight.type="sin", print.weighted.accuracy = TRUE)
wconfusionmatrix(m, weight.type="tanh", print.weighted.accuracy = TRUE)
wconfusionmatrix(m, weight.type= "custom", custom.weights = c(1,0.1,0),
                 print.weighted.accuracy = TRUE)