issue with the output

Question

Reen Blake-Carr 2020-8-17

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/580605-issue-with-the-output

评论： Walter Roberson 2020-8-18

Hi there i have been building a fuzzy inference system for the pima indians dataset which uses the triangular membership fucntions as seen below now they require three inputs as it is known and i have used these inputs from the basis of a decision tree i have created i am currently experiencing some kind of issue with the rule base now i have written the rules using a symbolic matrix format but i do keep getting this error when i try to use the 2nd rule using the eval function the error is when im trying to use the eval function for rule 2 it seems to jump to input 5 which i dont understand and i have tried everything to try to get it working but i am unable to and to which it gives me this warning to which there is a rule base and a range to which it should folllow and i have tried everything to try to get this working i am really new at this software and so im not very good at working the error codes and the code itself out yet so i am asking would anyone be able to help me solve these warnings as i need to get them done.

below is a excerpt of my code

any help given is grealty appreciated

Warning: Input 5 expects a value in range [0 1], but has a value of 2.

Warning: No rules fired for Output 1. Defuzzified output value set to its mean range value 0.5.

fis = mamfis('Name', "PimaIndians");

fis = addInput(fis,[0 160], 'Name',"Glucose");

fis = addmf(fis,"Glucose","trimf",[0 40 99],'Name',"low");

fis = addmf(fis,"Glucose","trimf",[99 110 127],'Name',"medium");

fis = addmf(fis,"Glucose","trimf",[127 140 157],'Name',"High");

fis = addInput(fis,[0 30],'Name',"BMI");

fis = addmf(fis,"BMI","trimf",[26 26.2 26.4],'Name',"low");

fis = addmf(fis,"BMI","trimf",[28 28.6 28.8],'Name',"medium");

fis = addmf(fis,"BMI","trimf",[29.3 29.6 29.9],'Name',"High");

fis = addInput(fis,[0 61],'Name',"BloodPressure");

fis = addmf(fis,"BloodPressure","trimf",[10 20 30],'Name',"low");

fis = addmf(fis,"BloodPressure","trimf",[30 40 45],'Name',"Medium");

fis = addmf(fis,"BloodPressure","trimf",[50 55 61],'Name',"High");

fis = addInput(fis,[0 30],'Name',"Age");

fis = addmf(fis,"Age","trimf",[0 5 10],'Name',"low");

fis = addmf(fis,"Age","trimf",[10 15 20],'Name',"medium");

fis = addmf(fis,"Age","trimf",[20 25 30],'Name',"High");

fis = addInput(fis,[0 1],'Name',"DiabetesFunction");

fis = addmf(fis,"DiabetesFunction","trimf",[0 0.1 0.2],'Name',"low");

fis = addmf(fis,"DiabetesFunction","trimf",[0.2 0.3 0.4],'Name',"medium");

fis = addmf(fis,"DiabetesFunction","trimf",[0.4 0.5 0.6],'Name',"High");

fis = addInput(fis,[0 6],'Name',"Pregnancies");

fis = addmf(fis,"Pregnancies","trimf",[0 1 2],'Name',"low");

fis = addmf(fis,"Pregnancies","trimf",[2 3 4],'Name',"medium");

fis = addmf(fis,"Pregnancies","trimf",[4 5 6],'Name',"high");

fis = addInput(fis,[0 30],'Name',"Insulin");

fis = addmf(fis,"Insulin","trimf",[0 5 10],'Name',"low");

fis = addmf(fis,"Insulin","trimf",[10 15 20],'Name',"medium");

fis = addmf(fis,"Insulin","trimf",[20 25 30],'Name',"High");

fis = addOutput(fis,[0 1],'Name',"Diabetes");

fis = addmf(fis,"Diabetes","trimf",[0 0.2 0.4],'Name',"NoDiabetes");

fis = addmf(fis,"Diabetes","trimf",[0.5 0.6 0.7],'Name',"HasDiabetes");

Rule1 = "Glucose==medium => Diabetes=NoDiabetes"

Rule2 = "BMI==low => Diabetes=NoDiabetes"

Rule3 = "BMI==High & Glucose==High => Diabetes=HasDiabetes"

Rule4 = "BloodPressure==low => Diabetes=NoDiabetes"

Rule5 = "BloodPressure==High & Age==low => Diabetes=NoDiabetes"

Rule6 = "BloodPressure==High & Age==High => Diabetes=HasDiabetes"

Rule7 = "Glucose==medium & BMI==low => Diabetes=NoDiabetes"

Rule8 = "BMI==low & Age==low => Diabetes=NoDiabetes"

Rule9 = "Age==medium & Glucose==low => Diabetes=NoDiabetes"

Rule10 = "Glucose==low & DiabetesFunction==low => Diabetes=NoDiabetes"

Rule11 = "DiabetesFunction==High & Pregnancies==high => Diabetes=HasDiabetes"

Rule12 = "Pregnancies==low & Insulin==low => Diabetes=NoDiabetes"

Rule13 = "Pregnancies==medium & Insulin==High => Diabetes=HasDiabetes"

Rules = [Rule1 Rule2 Rule3 Rule4 Rule5 Rule6 Rule7 Rule8 Rule9 Rule10 Rule11 Rule12 Rule13]

fis = addRule(fis,Rules)

evalfis(fis,[2])

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

Walter Roberson 2020-8-17

You need to pass inputs to evalfis not rule numbers

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Walter Roberson 2020-8-18

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/580605-issue-with-the-output#answer_481812

在 MATLAB Online 中打开

You have a fis with 7 inputs and 1 output. When you ask to

evalfis(fis, [2])

you are passing in one input instead of 7 inputs. That syntax does not ask that rule #2 be evaluated or displayed: it says that you want 2 to be the input for your entire system.

Now there are several different ways that evalfis could react when you pass in a scalar instead of the seven different inputs. The way that it happens to react is as if you had passed in the scalar for each of the inputs, so like

evalfis(fis, repmat(2,1,7))

However your 5th input, Diabetes Function, is only permitted to be in the range 0 to 1, so it ignores the 2 and treats it as half way through the valid range, 0.5

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

Walter Roberson 2020-8-18

There is no tool to exercise a particular fis rule given the rule number.

请先登录，再进行评论。

issue with the output

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

回答（1 个）

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

另请参阅

类别

标签

Community Treasure Hunt

issue with the output

1 个评论 显示 -1更早的评论隐藏 -1更早的评论

回答（1 个）

1 个评论 显示 -1更早的评论隐藏 -1更早的评论

另请参阅

类别

标签

Community Treasure Hunt

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

1 个评论
显示 -1更早的评论隐藏 -1更早的评论