While tuning the SVM classification model in Matlab, I came across the rng function in matlab in which seed (stabilizes the random shuffling of the data in the algorithm) is changed. When the function called is rng(1) then I am getting one accuracy value (99%). When it is changed to rng(2) then I am getting another value (57%). So there is a huge change in accuracy as visible. What does this mean? Am I training it wrong?
The train and test set correct rate (in %) that I am getting with different runs without changing rng are(train,test)
(96, 82.8)
(94.6, 95.3)
(96, 85.9)
(96, 90)
(95, 95)
$\begingroup$
$\endgroup$
2
1 Answer
$\begingroup$
$\endgroup$
4
The training errors in this dataset has a huge difference (99% vs 57%). So, maybe the one with the rng(1) split has overfitted your dataset.
So there is a huge change in accuracy as visible. What does this mean? Am I training it wrong?
The huge change might be due to overfitting. (Also, judge the model through validation curves, and then fit a model which balances the bias-variance plot.)
-
$\begingroup$ made a small change in the question. Let me know whether still there is over-fitting or not $\endgroup$– girl101Commented Aug 9, 2016 at 7:15
-
1$\begingroup$ @Rishika No, I don't think there's any overfitting there. But, that's without changing the rng value, right? $\endgroup$– Dawny33Commented Aug 9, 2016 at 7:17
-
-
1$\begingroup$ @Rishika The model looks good to me, then :) $\endgroup$– Dawny33Commented Aug 9, 2016 at 7:25
rng(1)is better thanrng(2)or maybe it is overfitting. $\endgroup$