Javierfdr
  • Member for 7 years
  • Last seen more than 1 year ago
  • Barcelona, Spain
1 answers
14 votes
33k views
Intuition for the regularization parameter in SVM
20 votes

The regularization parameter (lambda) serves as a degree of importance that is given to misclassifications. SVM pose a quadratic optimization problem that looks for maximizing the margin between both ...

View answer
3 answers
13 votes
13k views
How to choose a classifier after cross-validation?
Accepted answer
15 votes

You do cross-validation when you want to do any of these two things: Model Selection Error Estimation of a Model Model selection can come in different scenarios: Selecting one algorithm vs others ...

View answer
5 answers
7 votes
2k views
What would be a good way to use clustering for outlier detection?
6 votes

A very robust clustering algorithm against outliers is PFCM from Bezdek. In this paper Bezdek proposes Possibilistic-Fuzzy-C-Means which is an improvement of the different variations of fuzzy ...

View answer
2 answers
3 votes
1k views
Recommendation engine with mahout
3 votes

I recommend you to take a look to Oryx (https://github.com/OryxProject/oryx). Oryx is based on Apache Mahout (actually one of the creators of Mahout Sean Owen built it) and provides recommendation ...

View answer
2 answers
1 votes
208 views
Feature reduction convenience
Accepted answer
3 votes

Feature Selection (FS) methods are focused on specializing the data as much as possible to find accurate models for your problem. Some of the main issues that drive the need for FS are: Curse of ...

View answer
2 answers
5 votes
220 views
Algorithm to find common sequence
3 votes

Not a typical problem formulation indeed. I suggest two approaches: Use a simple bayesian model where the states or nodes are each of your features. Connect the nodes directly when they haven a ...

View answer
4 answers
3 votes
2k views
How to learn a classifier from a dataset with high imbalance
3 votes

A common strategy for dealing with imbalance is to penalize harder the missclassifications that select the class with higher frequency. In a binary classification problem you could penalize by ...

View answer
2 answers
1 votes
943 views
Can i find similar players using a clustering method like the k-mean algorithm?
2 votes

You can go as further or specific you want with this kind of data. I would suggest first to analyze the data thoroughly before deciding which algorithms to apply. Evaluate the mean, quartiles, max and ...

View answer
1 answers
1 votes
60 views
Which classifier is efficient in dealing with test query which belongs to no trained class?
2 votes

You have trained a model to recognize or discriminate between several particular classes. So when having a new test sample (that according to your knowledge) does not belong to any of this class, the ...

View answer
3 answers
10 votes
3k views
Log file analysis: extracting information part from value part
2 votes

This does not seem a Data Science problem. However there are very nice tools to do exactly that, checkout: logstash, flume and fluentd. Actually if you want to be able to filter in fast and "smart" ...

View answer
2 answers
1 votes
492 views
Neural Network Hidden Neuron Selection Strategy
2 votes

A rule of thumb approach is: start with a number of hidden neurons equal (or little higher) that the number of features. In your case it would be 9. My suggestion is to start with 9*2 = 18 to cover ...

View answer
4 answers
11 votes
10k views
Feature selection and classification accuracy relation
1 votes

About the specific question You should not expect a specific behavior (increase and then decrease of accuracy) while you select subset of features, since this will be totally dependent on the problem ...

View answer
1 answers
1 votes
173 views
In image classification/recognition how are images coded?
1 votes

This will depend on the type of task you want to perform: Object recognition is a wide task than can be approached in different ways such as: [Multi-part OR][1] [Multi-scale OR][2] Multi-class OR [...

View answer
1 answers
2 votes
147 views
Linear combination of weak estimators over fuzzy classifiers?
1 votes

You might find a solution for this by checking out Viola & Jones face detection algorithm (and object detection in general) http://www.cs.ubc.ca/~lowe/425/slides/13-ViolaJones.pdf. Particularly ...

View answer
1 answers
0 votes
203 views
What are the current killing machine learning methods?
Accepted answer
0 votes

The question is very general. However, there are some studies being conducted to test which algorithms perform relatively well in a broad range of problems (I'll add link to papers later), concerning ...

View answer