bogatron
  • Member for 7 years, 7 months
  • Last seen more than a month ago
Detecting cats visually by means of anomaly detection
Accepted answer
8 votes

You could simplify your problem significantly by using a motion/change detection approach. For example, you could compare each image/frame with one from an early time (e.g., a minute earlier), then ...

View answer
Cross-validation: K-fold vs Repeated random sub-sampling
Accepted answer
8 votes

If you have an adequate number of samples and want to use all the data, then k-fold cross-validation is the way to go. Having ~1,500 seems like a lot but whether it is adequate for k-fold cross-...

View answer
Which non-training classification methods are available?
Accepted answer
7 votes

What you are asking about is Instance-Based Learning. k-Nearest Neighbors (kNN) appears to be the most popular of these methods and is applicable to a wide variety of problem domains. Another general ...

View answer
What is a Dichotomy?
Accepted answer
6 votes

In a machine learning context, a dichotomy is simply a split of a set into two mutually exclusive subsets whose union is the original set. The point being made in your quoted text is that for four ...

View answer
Calculating entropies of attributes
5 votes

Consider the formula for Entropy: $E(S) = \sum_\limits{i=1}^{n}-p_{i}\log_2p_{i}$ Expanding that summation for the four concept decision attributes for your problem gives $E(S) = -p_{cinema}\log_2\...

View answer
Why do my Latent Dirichlet Allocation Topics mix words that never co-occurred?
4 votes

It could be partially due to the number of topics you selected but the fact that two words rank high for a given topic doesn't necessarily mean that the two words will frequently occur in the same ...

View answer
What is the difference between Latent and Explicit Semantic Analysis
4 votes

The difference is that with ESA, the concepts are already known and labeled (hence, "manifest concepts"), whereas in LSA the concepts are latent (they are undefined and need to be discovered). Note ...

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

For a complete decision tree, either of your proposed models would be able to represent the same set of concepts since a decision tree can be decomposed into a disjunction of conjunctions. I don't ...

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

K-means does not make an assumption regarding how many observations should be assigned to each cluster. K is simply the number of clusters one chooses to generate. During each iteration, each ...

View answer
Where does the sum of squared errors function in neural networks come from?
2 votes

Your reference of Bishop is not entirely accurate. What he states in the paper you linked is It should be noted that the standard sum-of-squares error, introduced here from a heuristic viewpoint, can ...

View answer
How can the performance of a neural network vary considerably without changing any parameters?
Accepted answer
2 votes

In general, there is no guarantee that ANNs such as a multi-layer Perceptron network will converge to the global minimum squared error (MSE) solution. The final state of the network can be heavily ...

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

There is nothing necessarily wrong with this. If you have no better information, then using past performance (i.e., prior probabilities) can work pretty well, particularly when your classes are very ...

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

[Since you've updated your question to refer to a different algorithm (changed k-means to kNN), I'm adding this as a separate answer specifically for kNN.] It appears you may still be confusing kNN ...

View answer
Group similarity by high dimension vector comparison
1 votes

You can't actually prove that the two groups are similar but you can establish a confidence level/threshold. Furthermore, it is possible that the two groups won't be similar (depending on your ...

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

It is possible to produce a corpus from the learned LDA parameters ($\theta$ and $\phi$) according to the generative model of LDA but it is not realistic to expect that you would recreate the original ...

View answer
How to find a confidence level given the z value
1 votes

To convert between z-scores and confidence values with python, use the cdf and ppf functions in scipy.stats.norm. There is a nice example of how to use them in the answer for this question.

View answer
Feauture selection for clustering regarding zero-correlated feature
Accepted answer
1 votes

Lack of correlation with other features is not a reason to omit a feature. On the contrary, it is usually a reason to keep the feature because it may provide unique information. Typically, highly ...

View answer
Neural networks with non-negative weights
1 votes

You can easily take an existing Multilayer Perceptron implementation and modify the backpropagation algorithm to prevent weights from becoming negative. Of course, you would also need to make sure you ...

View answer
How to appropriately set weights for weighted KNN
1 votes

A decaying exponential of the form $e^{-\alpha x}$ (where $x$ is the distance from the observation) is convenient in this situation. It has the nice feature that the weight is equal to $1$ when the ...

View answer
Why might several types of models give almost identical results?
1 votes

As @seanv507 suggested, the similar performance may simply be due to the data being best separated by a linear model. But in general, the statement that it is because the "observations to variable ...

View answer
Why are precision and recall used in the F1 score, rather than precision and NPV?
0 votes

If it is purely a binary classification problem (class A vs. class B), then the benefit of the F-score is primarily for characterizing performance over an unbalanced data set (more instances of one ...

View answer
Use test data as train: does it make sense?
0 votes

You haven't stated your reason for wanting to do this, nor what algorithm you are using, both of which will affect whether your proposed tactic "makes sense". I'll give you a couple reasons why the ...

View answer