For loop index is too large when learning with anfis

I have data 7×108 for input. And 7x1 for target. When i try to run it, it got "Warning about for loop index is too large. Truncating to 9223372036854775807".

    trnData = [input,target];
numMFs = 2;
mfType = 'gbellmf';
fismat = genfis1(trnData,numMFs,mfType);

How i can fix this problem, so i can make it learn with anfis? Thank you

[Walter Roberson]

What is your largest target (value) ? And is your target numeric?

[yuto nonoa]

This is my code for target:

target = zeros(7,1); target(1:7,:)=1;

[yuto nonoa]

Just wanna try it with 7 images with same gesture.

Answer 1

[Walter Roberson]

numMFs is the number of membership function per input. Your data has 108 columns so it has 108 inputs. Total number of membership functions is 2^108 .

You should be passing in the transpose of input -- and I do not think you should be including the target information with that.

[yuto nonoa]

My actual data performed for 7 different type of gesture. So i have 49 total images. And for target, i write it like:

 target = zeros(total_images,1);
 target(1:7,:) = 1;
 target(8:14,:) = 2;
 target(15:21,:) = 3;
 target(22:28,:) = 4;
 target(29:35,:) = 5;
 target(36:42,:) = 6;
 target(43:49,:) = 7;

Can you provide me code of numMFs with your solution?

Thanks for advance.

[Walter Roberson]

The problem is not that you specified 2 for the number of modular functions. The problem is that you passed in your input with rows and columns exchanged.

[yuto nonoa]

I already solved it with genfis2. Thank you @Walter Roberson.