0
$\begingroup$

I have a feature extraction function fe, which gives me n/0.816497 as the output of fe[{1}]:-

data = {{1}, {0}, {-1}};
fe = FeatureExtraction[data, {"NumericVector"}];
1/fe[{1}]

I want to modify fe and have say n/0.77 as the new output. How can I change it? I tried below codes, but have no idea how to put a back to Options[fe] to make the updates. Besides, below process to obtain/update the 0.816497 is clumsy, and feel free to let me know if there is any way to enhance.

a = Options[fe]
b = a[[1]];
c = b["Processor"];
d = c[[2]];
e = d["Processors"];
f = e[[4]];
g = f[[2]];
g["StandardDeviation"]

g["StandardDeviation"] = {0.77}
f[[2]] = g;
e[[4]] = f;
d["Processors"] = e;
c[[2]] = d;
b["Processor"] = c;
a[[1]] = b;
a
$\endgroup$
  • $\begingroup$ I don't really see the purpose in training a feature extractor when you want to manually specify the trained parameters... So I guess my question is: Why would you want to do this? $\endgroup$ – Lukas Lang Aug 12 '18 at 19:44
  • $\begingroup$ @Lukas Lang Sometimes when we input images, the feature extraction function give some specific number of features as output, say 11 or 12 feaures. I want a different number of features for further training, which can be half, double or even ten times of that number. I tried to modify that problem in another way but didn't got answer. Now I tried to simplify my question further on changing parameters only to try to get the first step done. $\endgroup$ – H42 Aug 13 '18 at 0:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.