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

learn more… | top users | synonyms (1)

4
votes
1answer
701 views

Rationale of using AUC?

Especially in the computer-science oriented side of the machine learning literature, AUC (area under the receiver operator characteristic curve) is a popular criterion for evaluating classifiers. ...
2
votes
2answers
190 views

Calculating the information contained in a message

Lets say i want to calculate the information content of a particular message.What apart from the message itself has to be taken into account in doing so, and what data would i need to collect to ...
5
votes
6answers
1k views

Testing for stability in a time-series

Is there a standard (or best) method for testing when a given time-series has stabilized? Some motivation I have a stochastic dynamic system that outputs a value $x_t$ at each time step $t \in ...
1
vote
1answer
223 views

Variable selection for increasing accuracy

I know that there are various posts regarding variable selection but I am asking something particular. With respect to the question that I posted today in the following link: Low accuracy in out of ...
4
votes
1answer
3k views

What is the difference between empirical variance and variance?

As far as I know variance is calculated as $$\text{variance} = \frac{(x-\text{mean})^2}{n}$$ while $$\text{Empirical Variance} = \frac{(x-\text{mean})^2}{n(n-1)} $$ Is it correct? Or is there ...
0
votes
0answers
183 views

How to calculate variance of three or more attributes with known frequencies

Question was originally posted on Stack Overflow. I want to calculate variance of data at a given time where I know the frequency of the conditional and decision attributes. At time $t$ every ...
20
votes
2answers
2k views

Sites for predictive modeling competitions

I participate in predictive modeling competitions on Kaggle, TunedIt, and CrowdAnalytix. I find that these sites are a good way to "work-out" for statistics/machine learning. Are there any other ...
4
votes
2answers
967 views

Training multiple models for classification using the same dataset

For my classification problem, I am trying to classify an object as Good or Bad. I have been able to create a good first classification step that separates the data into 2 groups using SVM. After ...
8
votes
2answers
275 views

Significance of initial transition probabilites in a hidden markov model

What are the benefits of giving certain initial values to transition probabilities in a Hidden Markov Model? Eventually system will learn them, so what is the point of giving values other than random ...
1
vote
3answers
388 views

How does a classifier handle unseen documents that do not belong to any of the pre-existing classes?

I am doing text classification, and have been playing around with different classifiers. However I have a pretty basic question: what if a new unseen document comes in and it happens to not belong to ...
11
votes
4answers
6k views

For classification with Random Forests in R, how should one adjust for imbalanced class sizes?

I am exploring different classification methods for a project I am working on, and am interested in trying Random Forests. I am trying to educate myself as I go along, and would appreciate any help ...
3
votes
2answers
716 views

Designing covariance matrix and kernel function for a gaussian process

I am not so experienced to design a customized covariance matrix / kernel functions. I would like to get such a understanding that after looking at data, I can figure out the covariance matrix. For ...
11
votes
1answer
862 views

Optimal algorithm for solving n-armed bandit problems?

I've read about a number of algorithms for solving n-armed bandit problems like $\epsilon$-greedy, softmax, and UCB1, but I'm having some trouble sorting through what approach is best for minimizing ...
5
votes
2answers
905 views

Questions about variable selection for classification, and different classification techniques

I have a question concerning feature selection and classification. I will be working with R. I should start by saying that I am not very familiar with data mining techniques, aside from a brief ...
2
votes
1answer
563 views

How to choose the 1st threshold/classifier/ weak learner in Adaboost?

I am having some difficulty understanding Adaboost. How should the 1st threshold/classifier/weak learner be chosen? It seems that there are two conditions which must be satisfied Choose the ...
6
votes
3answers
1k views

Naive Bayes classification for “That's what she said” problem

I became interested in doing this in C# for my own amusement after reading the following papers: http://www.cs.washington.edu/homes/brun/pubs/pubs/Kiddon11.pdf I also took a look at ...
6
votes
4answers
844 views

Number of features vs. number of observations

Are there any papers/books/ideas about the relationship between the number of features and the number of observations one needs to have to train a "robust" classifier? For example, assume I have 1000 ...
5
votes
1answer
319 views

What do “real values” refer to in supervised classification?

I'm using supervised classification algorithms from mlpy to classify things into two groups for a question-answering system. I don't really know how these algorithms work, but they seem to be doing ...
24
votes
6answers
2k views

Data mining: How should I go about finding the functional form?

I'm curious about repeatable procedures that can be used to discover the functional form of the function y = f(A, B, C) + error_term where my only input is a set of ...
1
vote
1answer
126 views

Optimize a boolean function

I have some data that a downstream system needs an optimized function of boolean logic for. Essentially I have data similar to: ...
6
votes
4answers
983 views

Improving accuracy of a binary classification when the target is unbalanced

I am working on the BRFSS dataset with the goal of predicting Diabetes. The dataset has 500,000 rows and 405 columns. It is a 0/1 classification problem, the ratio of 0 to 1 is 90:10. I tried using ...
7
votes
1answer
372 views

Information on how value of k in k-fold cross-validation affects resulting accuracies

I've been doing some Machine Learning, and have been using k-fold cross-validation to assess the generalisation performance of the algorithm. I've tried k-fold cross-validation with k = 5 and k = 200 ...
30
votes
3answers
675 views

Application of machine learning methods in StackExchange websites

I have a Machine Learning course this semester and the professor asked us to find a real-world problem and solve it by one of machine learning methods introduced in the class, as: Decision Trees ...
4
votes
1answer
456 views

Interpretation of a one cluster solution using the EM cluster algorithm

I'm trying to use the EM cluster algorithm, provided by the software Weka, to classify my data and it only finds one cluster. Could I interpret this as there are no ways to distinguish the ...
6
votes
2answers
246 views

Number of eigenfunctions for kernel

While studying machine learning, I've read the following statement: The kernel $K(x,y)=(x\cdot y+1)^d$ , for $x, y \in \mathbb{R}^p$, has $M={p+d \choose d}$ eigenfunctions that span the space of ...
1
vote
1answer
563 views

How to handle skewed binary target variables? [duplicate]

Possible Duplicate: Supervised learning with “rare” events, when rarity is due to the large number of counter-factual events I am trying to predict diabetes using the BRFSS ...
4
votes
2answers
274 views

Learning a univariate transform (kernel?) for novelty detection

I have 150 observations, 500 features, and I am interested in novelty detection (outlier detection): given a new observation (let's say 'patient') I want to know if it is different from the previous ...
12
votes
2answers
956 views

Supervised learning with “rare” events, when rarity is due to the large number of counter-factual events

Suppose you get to observe "matches" between buyers and sellers in a market. You also get to observe characteristics of both buyers and sellers which you would like to use to predict future matches ...
9
votes
1answer
1k views

Is there a way to explain a prediction from a random forest model?

Say I've got a predictive classification model based on a random forest (using the randomForest package in R). I'd like to set it up so that end-users can specify an item to generate a prediction for, ...
4
votes
1answer
356 views

Meaning and use of `reltol` in `nnet` library in R

I'm trying to use the nnet library in R, and can't seem to work out how to use the reltol parameter. It says in the docs: ...
3
votes
2answers
993 views

$\nu$-svm parameter selection

For the $\nu$-SVM (for both classification and regression cases) the $\nu \in (0;1)$ should be selected. The LIBSVM guide suggests to use grid search for identifying the optimal value of the $C$ ...
6
votes
2answers
2k views

How to get generalisation performance from nnet in R using k-fold cross-validation?

I'm doing some Machine Learning in R using the nnet package. I want to estimate the generalisation performance of my classifier by using k-fold cross-validation. ...
5
votes
2answers
618 views

Implementing the 'kernel trick' for a support vector machine in R

I've heard a bit about the 'kernel trick' for support vector machines, and I was wondering: How do you identify problems that might benefit from the kernel trick? How to implement it in R? Thank ...
5
votes
1answer
2k views

When would one want to use AdaBoost?

As I've heard of the AdaBoost classifier repeatedly mentioned at work, I wanted to get a better feel for how it works and when one might want to use it. I've gone ahead and read a number of papers and ...
3
votes
1answer
2k views

How do I use the GPML package for multi dimensional input?

I have downloaded the Gaussian Processes for Machine Learning (GPML) package (gpml-matlab-v3.1-2010-09-27.zip) from the website, and I can run the regression example (demoRegression) in Octave. It ...
5
votes
2answers
460 views

Deriving mathematical model of pLSA

After knowing how LSA works, I went on continue reading on pLSA but couldn't really make sense of the mathematical formula. This is what I get from wikipedia (other academic papers/tutorial show ...
0
votes
0answers
1k views

How to convert log likelihoods into scores in Naive Bayes?

I am currently implementing a text classification program with Naive Bayes. I produce two multinominal models in my training function: p(w|nonSPAM) and p(w|SPAM)) as well as a prior probability P(S). ...
1
vote
0answers
270 views

Confusion in MLE and EM [closed]

I was trying to read through Maximum Likelihood Estimation(MLE) and Expectation and Maximization(EM) algorithm. But while reading them, I got two interpretations. I am trying to post my questions, ...
41
votes
7answers
11k views

Machine Learning using Python

I am considering using Python libraries for doing my Machine Learning experiments. Thus far, I had been relying on WEKA but have been pretty dissatisfied on the whole. This is primarily because I have ...
2
votes
1answer
200 views

Interpretation of “one” feature change in a supervised classifier

i'm making experiments using app. 5000 labeled dataset.i'm trying different supervised ML algorithm to evaluate the results.The vector size is 13 with the labels (totally 12 features+1 label) and i ...
2
votes
2answers
619 views

KL divergence calculation

I am wondering that how one can calculate KL-divergence on two probability distributions. For example, if we have ...
20
votes
2answers
797 views

How to choose between learning algorithms

I need to implement a program that will classify records into 2 categories (true/false) based on some training data, and I was wondering at which algorithm/methodology I should be looking at. There ...
6
votes
2answers
510 views

Overfit by removing misclassified objects?

Actually this question may be simple for you, but I need to learn the correct answer. If I remove misclassified instances from data set with Naive Bayes (it gives minimum FP rate) and then train ...
3
votes
2answers
278 views

Resources about Gibbs sampling in hybrid Bayesian networks

I've been trying to get my hands on a substantial resource for using Gibbs sampling in hybrid Bayesian networks, that is, networks with both continuous and discrete variables. So far I can't say I ...
2
votes
1answer
1k views

Bayesian classifier with multivariate normal densities

Supposing a Bayesian classifier with multivariate normal densities, how do I find the error rate of the classifier when we have two classes? I am using this: When dimension $d = 1$: $$P(x | \mu , ...
3
votes
1answer
283 views

Help with deploying a model

The two main packages I use at work are Palisade Risk and SPSS Clementine - they are both quite old versions and I've been supplementing the ability to analyze properly at work with more modern ...
3
votes
2answers
1k views

Advice on classifier input correlation

I have a lot of data on previous race history and I'm trying to predict a percentage chance of winning the next race using Regression, kNN, and SVM learning algorithms. Say a race has 5 runners, and ...
1
vote
1answer
376 views

Bayes classifier

Given the prior probability of 2 distributions, $N(x,y)$ and $N(a,b)$, where $N(\mu,\sigma^2)$: How do you make a decision rule to minimize the probability of error, if the prior probabilities are ...
1
vote
0answers
204 views

The relationship between machine learning, data mining and statistical analysis? [duplicate]

Possible Duplicate: What is the difference between data mining, statistics, machine learning and AI? How to compare machine learning vs. data mining? data mining vs. statistical analysis? ...
4
votes
1answer
231 views

Hidden states in hidden conditional random fields

I am trying to study hidden conditional random fields but I still have some fundamental questions about those methods. I would be immensely grateful if someone could provide some clarification over ...