Methods and principles of building "computer systems that automatically improve with experience."

learn more… | top users | synonyms

0
votes
1answer
10 views

What's the best algorithm type for low-dimensional grouping

I'm looking for some advice on directions to head in a project I'm working on. Basically what I want to do is identify general (of varying size) groups in a 2-D grid of points belonging to one of ...
1
vote
1answer
19 views

Understanding differences between large and small dimensional data when implementing algorithms

I'm working on a local outlier factor implementation based on the wikipedia entry : http://en.wikipedia.org/wiki/Local_outlier_factor This article seems to explain it in just two dimensional data. ...
0
votes
1answer
10 views

Question of unary term (data term) of the graph cut method

I am trying to apply graph cut method for my segmentation task. I found some example codes at Graph_Cut_Demo. Part of the codes are showing below ...
0
votes
0answers
19 views

predicting time series with support vector machine using R

I am planning to do time series prediction using support vector Machine. I could not find any materials about time series application of support vector machines using R or Mat-lab. Similar question ...
0
votes
1answer
17 views

In a neural network of n features which predict a continuous variable X, how to tell the feature which contributes the most to the output value X?

Say I have a neural network which uses some input features say N, some inut layers say L which predict a continous variable say X. Can we say which features or combination of 2 features of the initial ...
0
votes
2answers
25 views

Varying LIBSVM predictions based on test series labels

So I have a pretty well testing SVC train series which puts me into the mid 80 percentile without outrageous C/g values. My current C value is 2.0 and gamma is 0.5. Good numbers across the range ...
0
votes
0answers
16 views

How to prepare my data for SVM classifier in matlab

I am new to SVM and Matlab. I would like to have an example how to prepare my data to be as input to the SVM classifer (using libsvm) let us assume that i have a group of words first i have ...
0
votes
0answers
15 views

Combined Two CLassifiers

I am involved in a research where i need to classify group of words (strings) into two classes I am currently reached a dead point where my classifier is not doing as i expected. I used like three of ...
0
votes
0answers
43 views

Random variables of mixed models

I am thinking about using mixed models as part of my research, but I am having trouble understanding its application. In particular, I have two somewhat related questions regarding mixed models. ...
0
votes
1answer
9 views

Can adaboost choose the same variable for multiple splits for a given tree?

Can adaboost choose the same variable for multiple splits for a given tree? The model was given 100 + variable to choose from and it did choose them for the other trees in the ensemble. I am using ...
0
votes
0answers
11 views

Naive based model for positive negative classes

I am new to the concept of machine learning and I am trying to figure out this problem. I have 4 clusters who's means kind of form a square. The training dataset that I provide to Naive Bayes has 2 ...
0
votes
0answers
17 views

Live selection of movie to suggest based on similarity of users

I am working with movie selection for users. 1 ) One of the first ways I thought was taking all the clicked only movie data and building decision trees out of it. Then when input is passed, the ...
0
votes
0answers
13 views

get a list of predicted variables based on rank

I have a data of device, location, age etc of some users, along with keywords that they clicked (or not). Based on this I want to build a model that predicts a list of keywords for test set that ...
2
votes
1answer
16 views

Overfitting with low test error

I have an ensemble of models that yields overfitting by some models by looking at the differentce between training and test error and also overfits in the combined ensemble of models providing a very ...
0
votes
0answers
14 views

Extracting fixed-length feature vectors from variable-length time series

I have a classification problem where I would like to develop a binary classifier to classify between two different types of objects, given a time-series (signal) related to that object. The problem ...
0
votes
1answer
14 views

Requirements for a valid neural network activation function?

What rules define a valid neural network activation function, excluding biological plausibility? What set of principles do softmax, rectified linear units, hyperbolic tangent, sigmoid, etc. follow? ...
2
votes
1answer
37 views

How to find all optima in an optimization problem?

I have an optimization problem where several optima can exist at different input values, and I need to find as many as possible. As an example consider the cross-in-tray function, which has four ...
2
votes
0answers
42 views

What enforces features diversity in RBM?

I'm working on an implementation of a Restricted Boltzman Machine (RBM). I made some tests on the MNIST dataset trying to learn a representation of the digit 2. My inputs are binary images. My aim is ...
0
votes
0answers
43 views
+50

Suggestions needed about classifier fusion

I'm working on a classification problem which involves two classifier to observe a single event. I'm providing a high level description of the problem without going into the technical details (the ...
0
votes
0answers
8 views

Partially sparse vectors for training classifiers

Is it a bad idea to use a partially sparse vector for training a logistic regression classifier? By "partially sparse", I mean that about half the vector is actually dense, with real valued numbers ...
0
votes
1answer
50 views

Machine learning - curve categorisation

I have curves of the following structure (it is the blue one I am interested in) These curves reflect the volume of blood (actually gamma ray counts) in the left ventricle as a function of time ...
0
votes
1answer
28 views

Statistical Commute Analysis in Java

I have a rather large commute every day - it ranges between about an hour and about an hour and half of driving. I have been tracking my driving times, and want to continue to do so. I am capturing ...
0
votes
0answers
69 views

How to best make millions of forecasts using time series data?

I need to make roughly 50 million forecasts every night. The data is daily, hierarchical (~50 million base series), intermittent/sparse (for many of the time series, lots of days have 0's), and not ...
0
votes
0answers
16 views

Feature/Variable selection to accompany mixed models?

I am trying to conduct an exploratory/data mining analysis to discover what socioeconomic factors best predict grade-school performance in children. I have a dataset with about 50000 ...
2
votes
1answer
52 views
+100

Optimal classification model for translating words

I have the following problem: I have a set of English words which I want to translate to Dutch. Of each words I mined a set of possible translations. For example, for the word "Eighteen" I obtained ...
1
vote
1answer
41 views

Algorithm to find subsets with high correlation

I have a reasonably large dataset (d) with predictor variables x1...xn and a target variable y. I can use recursive partitioning (such as CART or rpart in R) to find subsets of d with a high (or low) ...
0
votes
0answers
7 views

Is it safe to mix different encodings of features in one data set

I'm investigating which feature encoding yields the best results for training a predictive model on a biological problem. All variables have the same format and possible values before conversion to ...
0
votes
0answers
7 views

VC-Dimension of n-node binary decision tree in N-dimension feature space

Given input feature space $\mathcal{X} =\{0, 1\}^N$ and output label space $\mathcal{Y}=\{0,1\}$ , prove that the VC-dimension of a binary decision tree with $n$ nodes is in $O(n\text{log}N)$. I've ...
1
vote
1answer
29 views

Working with few data examples

I have been asked often in some interview, that how we should proceed when we have less data examples(say 50 or 100). What considerations needs to be made while choosing any algorithm. few points ...
0
votes
1answer
13 views

How to improve the performance of K-nn algorithm in R?

I am having a digit recognizer data set which has column names as label, pixel0, pixel1...pixel783. pixel values vary from 0 to 255 indicating the lightness or darkness of that pixel, with higher ...
0
votes
0answers
24 views

How do the various distributions work into machine learning

I'm a student who recently started taking a machine learning course and I'm trying to understand how the various distributions fit into the whole thing. From the research I've been doing it seems the ...
0
votes
0answers
18 views

$\chi^2$ test vs F-test in feature selection

In the context of feature selection for classification, does it make sense to use one filter based on $\chi^2$ test and the other one based on F-test? Or they are "interchangeable"?
0
votes
0answers
9 views

cross validation for kmodes in r

I am using k-modes (link) from the KlaR library (link) to cluster text data. I am not sure how to determine predictive error and thus perform cross-validation. Here is the "toy" sample, lets use ...
2
votes
0answers
47 views

How to split a decision tree when information gains of all attributes are zero?

The textbook tells us that we should choose an attribute with the maximum information gain to split a decision tree. My question is what if all information gains are zero? Should we stop splitting or ...
0
votes
0answers
18 views

Numerical Problems in Mixture of Gaussians Classifications

I am doing two-class classification with Gaussian Mixture Models (GMMs). If I understand it correctly I have to build two models $p(x | C1)$ and $p(x | C1)$ for the probability of input $x$ given ...
0
votes
0answers
19 views

Best classifiers for large data sets?

I'm working on a data set that contains electricity consumption data. There will be 2-3 features used. I'm not sure if that is all of the features to be used. Also, it will be a really large data set. ...
1
vote
0answers
9 views

How do i generate variables that are relevant only for some classes?

I want to generate data for classification. I've generated data with 10 variables with two are relevant for all classes and 8 noise. now, I want to generate variables that are relevant just for some ...
0
votes
0answers
37 views

How to represent outliers for multi dimensional data (local outlier factor)

Below graph taken from http://en.wikipedia.org/wiki/Local_outlier_factor displays "LOF scores : LOF image : This is great for two dimensional data but what about data > than two dimensions. How ...
1
vote
2answers
29 views

Newbie: Decision Tree in R with Two classes(Yes/No) where one class (No) is much larger than other class (Yes)

I am trying to make a decision tree using 4 features (A,B,C,D) to predict an out come for two classes E(Yes, NO). The problem is that the number of observations in my dataset that belong to one class ...
0
votes
0answers
5 views

Sample weights for classification problems

How can certain samples in the training set be prioritized (given more weights) in classification problems? What is the formal methodology to do so?
2
votes
1answer
25 views

What kind of functions can have non whole degrees?

Thanks for the help in advance. I am reading a technical report on a regression algorithm that reports a pair of functions as having a total degree of freedom of 5.4. I believe that both of these ...
0
votes
1answer
39 views

Selecting most realistic C and g params after gridsearch

I just ran an extended SVC gridsearch in libsvm on about 9000 multi-dimensional vectors representing a time series. Here are the highest scoring results: ...
1
vote
1answer
47 views

How to evaluate a clustering/unsupervised learning problem with massive amounts of data, with labels only for a small fraction of points

I'm wondering if anybody can point me to work on the evaluation of unsupervised learning where there are a very large (say hundreds of millions) number of points and manual labelling can only ever be ...
0
votes
1answer
32 views

How to prove that the manifold assumption is correct?

In machine learning, it is often assumed that a data set lies on a smooth low-dimensional manifold (the manifold assumption), but is there any way to prove that assuming certain conditions are ...
0
votes
0answers
9 views

Unsupervised feature learning from raw text as a previous step for clasification?

I have a corpus of 2500 opinions, is it posible to use scikit´s restricted boltzmann machine implementation to extract a feature vector as a previous step to a classification task?. What aproach do i ...
1
vote
0answers
20 views

Hard Case - prediction of chain stores revenue

Data about average monthly revenue from 2000 stores around whole country. Gini coeff. of reve around 20%, with 50% of observation around average, very thin tails of distribution Explanatory ...
2
votes
0answers
17 views

mob model tree algorithm

I am trying to figure out the inner workings of the mob function in the party package. I can't figure out how the splitting variable is selected when it is a categorical variable. In the publications ...
0
votes
1answer
27 views

About grid search to find the best value of C

I know when I want to find the best values of C and gamma, I should use grid-search. But in my case I want to find just the best value of C. So, this is called a line-search. Is there any function ...
0
votes
0answers
39 views

Logistic Regression, SVM or NN?

Just attended Andrew Ng’s online course on ML and although I’ve understood the methods I seem to be missing the intuition on where to apply them in terms of classification problems. What are the ...
0
votes
0answers
28 views

How to handle systematically missing values?

In my situation, one of two sources is not invoked if the confidence reported by the first source is higher than a threshold and hence it is missing in some examples. How can account for such missing ...