reset password
Author Message
skim144
Posts: 63
Posted 12:57 Feb 25, 2016 |

Hey guys, I got a really low accuracy for SVM (all default parameters), so I wanted to hear what everybody got. Here is mine:

Naive Bayes: 0.80

SVM : 0.59

Logistic Regression: 0.85

Random Forest: 0.88

steveshim
Posts: 6
Posted 11:39 Feb 26, 2016 |

Kind of changes each time I run it, but roughly:

NB: 0.82

SVM: 0.82

LR: 0.83

RF: 0.85

rkmx52
Posts: 23
Posted 23:19 Feb 26, 2016 |

I have:

NB: 0.80

SVM: 0.76

Logistic Regression: 0.81

Random Forest: 0.85

lmann2
Posts: 156
Posted 12:42 Feb 27, 2016 |

I have wildly different results, lol.  What columns did you choose are your y labels (I choose the last column)?  What did you choose are your test size (mine is .3)? What functions are you using for accuracy (I'm using metrics.accuracy_score).

NB: .66

SVM: .88888

LR: .89

Forest: .903

Last edited by lmann2 at 12:43 Feb 27, 2016.