News

FuzzyR 2.3

New features

• Add implementation of TSK type fuzzy inference systems based on An extended ANFIS architecture.
• Add new function tipper.tsk.

Bug fixes

• Address the issue of evalfis for interval type-2 fis: it now gives a single (rather than two) crisp defuzzified output.

FuzzyR 2.2

New features

• Add err.trn.fix flag to function anfis.optimise.
• Add height option to membership functions: gbellmf, gbell.fuzzification, it2gbellmf, singletonmf, gaussmf, it2gaussmf, trapmf, it2trapmf, trimf, it2trimf, trimf.fuzzification.
• New functions: gauss.fuzzification, it2gbell.fuzzification.
• Change functions addfis, addmf, evalfis to include interval type-2 features (Mamdani).

Bug fixes

• Change function evalfis to address the issue of a fis with multiple outputs.
• Address the issue of initialisation when a == 0 for funtions init.params.gbell and init.params.it2gbell.

Other changes

• Export function: writefis.
• Add DOI info for function km.da.
• Add new package dependency: grid.
• Add reference: A Comment on “A Direct Approach for Determining the Switch Points in the Karnik–Mendel Algorithm”.
• Add reference: Type-1 and Interval Type-2 ANFIS: A Comparison.

FuzzyR 2.1

New features

• Add Interval Type-2 ANFIS.
• Add accuracy measure UMBRAE: A new accuracy measure based on bounded relative error for time series forecasting.
• Add functions: fuzzyr.match.fun, fuzzyr.accuracy, km.da, it2gbellmf, it2gbell.fuzzification.

Bug fixes

• Customise function match.fun to address the issue that functions cannot be matched due to environmental settings.
• Change function gensurf to address a color issue.
• Change function defuzz to address an issue when all membership grades are equal to zero.
• Change function anfis.dMF.dP.gbellmf to address the issue of infinite outputs due to large factors.
• Change function plotmf to address an issue of customised label.

Other changes

• Hide functions: anfis.addnode, anfis.optimise.lse.
• Minor changes to functions: anfis.optimise, anfis.optimise.gradient, anfis.optimise.lse, anfis.dO3.dO2, anfis.plotmf, plotvar, plot_graph.
• Change function anfis.L2.which to use the direct approach for determining the switch points.
• Add reference: An extended ANFIS architecture and its learning properties for type-1 and interval type-2 models