User bogatron

8 votes

Accepted

Cross-validation: K-fold vs Repeated random sub-sampling

answered Jun 20, 2014 at 19:15

8 votes

Accepted

Detecting cats visually by means of anomaly detection

answered Jun 24, 2014 at 13:34

7 votes

Accepted

Which non-training classification methods are available?

answered Jul 2, 2014 at 14:14

6 votes

Calculating entropies of attributes

answered Feb 11, 2015 at 18:32

6 votes

Accepted

What is a Dichotomy?

answered Feb 11, 2016 at 19:51

4 votes

What is the difference between Latent and Explicit Semantic Analysis

answered May 14, 2015 at 13:41

4 votes

Why do my Latent Dirichlet Allocation Topics mix words that never co-occurred?

answered Aug 3, 2015 at 13:31

3 votes

What's the best way to use binned data in a tree-based model?

answered Feb 9, 2016 at 21:35

3 votes

kNN - what happens if more than K observation have the same distance to the centroid of the cluster

answered Dec 9, 2014 at 21:33

2 votes

kNN - what happens if more than K observation have the same distance to the centroid of the cluster

answered Dec 10, 2014 at 13:54

2 votes

Can distribution values of a target variable be used as features in cross-validation?

answered Jan 26, 2015 at 19:24

2 votes

Where does the sum of squared errors function in neural networks come from?

answered May 22, 2015 at 14:01

2 votes

Accepted

How can the performance of a neural network vary considerably without changing any parameters?

answered Mar 12, 2015 at 13:18

1 vote

Neural networks with non-negative weights

answered May 6, 2015 at 13:33

1 vote

Accepted

In Latent Dirichlet Allocation (LDA), is it reasonable to reconstruct the original bag-of-words using the document and word representations?

answered Mar 16, 2016 at 19:13

1 vote

How to appropriately set weights for weighted KNN

answered Feb 19, 2015 at 14:59

1 vote

Why might several types of models give almost identical results?

answered Aug 18, 2014 at 17:25

1 vote

Group similarity by high dimension vector comparison

answered Apr 21, 2016 at 13:53

1 vote

Accepted

Feauture selection for clustering regarding zero-correlated feature

answered Jan 12, 2016 at 14:41

1 vote

How to find a confidence level given the z value

answered Feb 5, 2016 at 15:23

0 votes

Use test data as train: does it make sense?

answered Mar 28, 2016 at 20:45

0 votes

Why are precision and recall used in the F1 score, rather than precision and NPV?

answered Apr 4, 2016 at 17:34

Stack Exchange Network

22 Answers

Cross-validation: K-fold vs Repeated random sub-sampling

Detecting cats visually by means of anomaly detection

Which non-training classification methods are available?

Calculating entropies of attributes

What is a Dichotomy?

What is the difference between Latent and Explicit Semantic Analysis

Why do my Latent Dirichlet Allocation Topics mix words that never co-occurred?

What's the best way to use binned data in a tree-based model?

kNN - what happens if more than K observation have the same distance to the centroid of the cluster

kNN - what happens if more than K observation have the same distance to the centroid of the cluster

Can distribution values of a target variable be used as features in cross-validation?

Where does the sum of squared errors function in neural networks come from?

How can the performance of a neural network vary considerably without changing any parameters?

Neural networks with non-negative weights

In Latent Dirichlet Allocation (LDA), is it reasonable to reconstruct the original bag-of-words using the document and word representations?

How to appropriately set weights for weighted KNN

Why might several types of models give almost identical results?

Group similarity by high dimension vector comparison

Feauture selection for clustering regarding zero-correlated feature

How to find a confidence level given the z value

Use test data as train: does it make sense?

Why are precision and recall used in the F1 score, rather than precision and NPV?