Description Usage Arguments Value References Examples
Calculates common classification accuracies measures known from remote sensing applications based on a Confusion Matrix that compares Reference classes to their respective class predictions. The computed accuracy measures are:
Overall Classification Accuracy (OAA)
Producer's Accuracy (PA)
User's Accuracy (UA)
Cohen's Kappa Coefficient (k)
Quantity Disagreement
Allocation Disagreement
1 2 3 4 5 6 7 8 | classAccuracy(..., conf.level = 0.95)
## S3 method for class 'hsmclass'
classAccuracy(object, conf.level = 0.95, ...)
## Default S3 method:
classAccuracy(refdata, predictions, equal.int = NA,
def.int = NA, conf.level = 0.95, ...)
|
... |
further arguments passed to or used by methods. |
conf.level |
the confidence level used to compute the confidence intervals of the overall accuracy. |
object |
object of class |
refdata |
|
predictions |
|
equal.int |
an equidistant class interval to be evaluated as classifciation scheme. Defaults to |
def.int |
a |
classAccuracy
returns an object of class "classaccur"
An object of class "classaccur"
returns a list
of the following components:
rsquared_val |
the coefficient of determination of the prediction model, calulated based on
|
.
predictions |
|
classwidth |
|
def.classbreaks |
|
equal.classbreaks |
|
overall.accuracy |
the overall accuracy of the classification scheme |
conf.oaa |
|
prodaccuracy |
a |
usersaccuracy |
a |
no.ref.classes |
a |
cohenskappa |
Cohen's Kappa Coefficient |
map.accuracy |
the map accuracy |
Quantity.Disagreement |
the quantity disagreement |
Allocation.Disagreement |
the allocation disagreement |
Congalton, R.G.; Green, K. Assessing the Accuracy of Remotely Sensed Data: Principles and Practices; Lewis Publications: Boca Raton, FL, USA, 1999; p. 137.
Richards, J.A. Remote Sensing Digital Image Analysis: An Introduction, 5th ed.; Springer: Berlin, Germany, 2013.
Hill, A., Breschan, J., & Mandallaz, D. (2014). Accuracy assessment of timber volume maps using forest inventory data and LiDAR canopy height models. Forests, 5(9), 2253-2275.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 | #----
# 1.) Example: classification accuracy for equidistant class width of 100:
#----
acc.equal<- classAccuracy(refdata.gr, predictions.gr, equal.int = 100)
summary(acc.equal)
#----
# 2.) Example: classification accuracy for arbitrary class breaks:
#----
acc.def<- classAccuracy(refdata.gr, predictions.gr,
def.int = c(0, 150, 200, 430, 610, 880))
summary(acc.def)
#----
# 3.) Example: classification accuracy for optimal class breaks:
#----
# run HSMclass:
## Not run:
hsm<- HSMclass(refdata.gr, predictions.gr, nclasses = 6,
iterations = 1000, coolfactor=0.99, InitTemp = 80,
weight.norefs = 2, weight.classwidth = 2)
# calculate accuracy:
acc.opti<- classAccuracy(hsm)
summary(acc.opti)
## End(Not run)
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.