Reputation
15,684
Next tag badge:
86/100 score
21/20 answers
Badges
7 44 93
Newest
 Yearling
Impact
~1.4m people reached

Jul
28
comment Should parsimony really still be the gold standard?
@CagdasOzgenc For instance a large random subspace ensemble.
Jul
28
comment Should parsimony really still be the gold standard?
On the other hand, when you have millions of variables and few objects, it is likely that purely by chance some variables are better at explaining outcome that the true interaction. In such case parsimony-based modelling will be more susceptible to overfitting than a brute-force approach.
Jul
20
revised Hidden Markov model - formulas
edited body; edited title
Jul
19
awarded  Yearling
Jul
13
awarded  Great Answer
Jul
7
awarded  Enlightened
Jul
7
awarded  Nice Answer
Jun
30
comment Feature selection with genetic algorithm in R
Still, I doubt it will work, unless you have tons of cases; so please proceed with caution and validate well.
Jun
28
revised Feature selection with genetic algorithm in R
deleted 1 character in body; edited title
Jun
28
answered Feature selection with genetic algorithm in R
Jun
28
revised Decreasing training learning curve
edited title
Jun
28
revised Three-ways Kruskal-Wallis test
Formatting.
Jun
18
awarded  Popular Question
Jun
17
comment Evaluation and optimization in machine learning
Also, as both evaluation and optimisation require some truth (i.e. true decision for some objects), they musn't ever use the same information, because it will make evaluation result biased and may hide overfitting.
Jun
9
comment How does extreme random forest differ from random forest?
@smci Fixed, thanks.
Jun
9
revised How does extreme random forest differ from random forest?
deleted 12 characters in body
Jun
8
answered Area under ROC curve for random forest
Jun
8
revised Area under ROC curve for random forest
added 31 characters in body
Jun
7
awarded  Nice Answer
Jun
4
awarded  Nice Answer