Tagged Questions

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

learn more… | top users | synonyms (1)

0
votes
0answers
15 views

Is There A Machine Learning Algorithm For Textual Data With Thousands Of Classifiers?

I've been asked to migrate this from StackOverflow to CrossValidated. I have a problem that I think Machine Learning can solve but am having a very hard time determining which ML Algorithm to use and ...
0
votes
0answers
15 views

Chi square and zscore - chose which one?

I'm newbie in machine learning. Recently I tried to learn something on this and got following concern: I have products classed by categories. Also I have users with gender and device model ...
0
votes
0answers
21 views

Advice for feature selection or feature extraction with semi-supervised learning

I am trying to solve a semi-supervised learning problem using LaplacianSVM. However, before applying LapSVM I would like either to perform feature selection or feature extraction. Furthermore, after ...
0
votes
0answers
17 views

Handling Sparse Data Frames - algorithm selection

I am new to machine learning/statistical modelling. I am trying to run a classification on a highly sparse dataset with 100 features, most of which are categorical (TRUE/FALSE) with the remaining ...
0
votes
2answers
21 views

k-mean clustering of week-times

I have data of meeting times. The data has weekday and hour of the day. I want to cluster the meeting times (I have reason to believe there are two different kinds of meetings that tend to occur at ...
1
vote
0answers
10 views

Data At Varying Granularity

I'm sorry for asking such a simple question, but for some reason it is throwing me off. By "granularity" I mean level of the data. For example, say in the classic example of spam classification you ...
1
vote
0answers
27 views

why do offline learning algorithms perform better than their online learning counterparts?

(I'm assuming infinite data, finite time for this comparison) I was wondering why it is exactly that online learning algorithms usually perform more poorly than their offline counter-parts.. Does ...
0
votes
0answers
12 views

Analysis of Feature Importances when features are dependent on one another

I can use random forests to determine which features are important when doing a prediction problem; for example. < height, weight, IQ measure> -> Is considered obese? Applying random forests ...
0
votes
1answer
20 views

Multi-class Classification using SVM with PCA

I'm doing an image classification task and the number of features of each example image is pretty huge (3,072: # pixels in each image). I'm thinking of using PCA to reduce the # features of each image ...
1
vote
0answers
25 views

How to estimate the correlated individual components from a sum, for a random process?

Assume that there are $N$ realisations of five individual, random variables$X_1$, $X_2$, $X_3$, $X_4$ and $X_5$, which in general could be correlated. We define another random variable ...
5
votes
1answer
31 views

Understanding sample complexity in the context of uniform convergence

I was reading Andrew Ng's notes and on page 6 he mentions (uniform convergence): $$Pr[\forall h \in \mathcal{H}_{finite}|\epsilon(h_i)-\hat{\epsilon}(h_j)| \leq \gamma] \geq 1-2ke^{-2\gamma^2m}$$ ...
0
votes
0answers
14 views

In Hidden Markov Model (HMM), is the transition matrix known, inferred, or assumed?

I'm reading Kevin Murphy's Probabilistic Machine Learning, which explains the forward algorithm to do filtering in HMM as follows (pp 610): The very first line says that the transition matrices ...
2
votes
1answer
22 views

Number of variables for decision trees

I have a data with just 5 independent variables and a response. I am dealing with a classification problem. Will decision trees perform well or the number of variables have to be higher to get ...
1
vote
1answer
36 views

Combining pca and classification algorithms

For some classification algorithms, assuming independence of data helps reduce the number of parameters to estimate. Why then not just to apply a method like pca or ica to the original features to get ...
0
votes
1answer
23 views

Machine learning : learn feature value range for a classification

Which domain the problem belongs to? Given a set of products some are classified as cheap and some not. The task is to determine the price range (probablistic) for cheap products ? Supervised ...
0
votes
1answer
34 views

Machine Learning for Text Classification

I am new to Machine Learning.I am working on a project where the machine learning concept need to be applied. Problem Statement: I have large number(say 3000)key words.These need to be classified ...
0
votes
0answers
30 views

Unsupervised learning algorithems to detect anomaly in waves

I have a sample of graphs (more then 10000...). that look like in the image below: I am searching an Unsupervised learning algorithems thet can help me to detect Anomaly observations. Here what i ...
0
votes
0answers
20 views

How can one go about recognizing a kind of motion using 3D depth data?

I'm using a Kinect device, and I'm currently extracting Joints, and Depth data unto probably a buffer data of 15 frames. This is done at 30 frames per second. The whole point of it is to try and ...
0
votes
0answers
26 views

Making a neural network model more sensitive to one of its several inputs

I am currently using neural network methods in R to model energy consumption (response) based on temperatures, previous consumption values and weekend dummy variables (inputs). Unfortunately, the ...
0
votes
0answers
24 views

SVM Dual Formulation :: KKT Constraint

In Andew Ng's SVM course notes, the final hard margin optimization problem is given as the following: I am unclear how to see from this where the 5th constraint is satisfied. The definition of ...
0
votes
0answers
10 views

Support vector regression in weka

I am using SVR for statistical down-scaling of precipitation. I have taken the first 3 factor scores in principal component analysis of variables as predictors and precipitation as predictand. As ...
0
votes
0answers
32 views

In natural language processing (NLP), how do you make an efficient dimension reduction?

In NLP, it's always the case that the dimension of the features are very huge. For example, for one project at hand, the dimension of features is almost 20 thousands (p = 20,000), and each feature is ...
0
votes
0answers
26 views

Gradients of marginal likelihood of Gaussian Process with squared exponential covariance, for learning hyper-parameters

The derivation of gradient of the marginal likelihood is given in this pdf, equation 5.9. But the gradient for the most commonly used covariance function, squared exponential covariance, is not ...
2
votes
1answer
69 views

How to estimate the individual components from a sum for a random process?

We have $N$ realisations of five individual, IID random variables $X_1$, $X_2$, $X_3$, $X_4$ and $X_5$. We define another random variable $S = X_1+X_2+X_3+X_4+X_5$. Now, for a given $S$ generated from ...
2
votes
0answers
25 views

Connections between d' (d-prime) and AUC (Area Under the ROC Curve); underlying assumptions

In machine learning we may use the area under the ROC curve (often abbreviated AUC, or AUROC) to summarise how well a system can discriminate between two categories. In signal detection theory often ...
3
votes
2answers
44 views

Mining patterns in continuous sequence

I have data in form of $N$ sequences $s_j=(t_i, e_i)_{i\in\{1,\ldots,n_j\}}$ with $n_j$ data-points each, where $t_i$ is a time-stamp and $e_i$ is a categorial event, say $e_i\in\{A,B,C,D\}$. The $N$ ...
0
votes
0answers
9 views

Is it OK to resample data in one vs all multiclass classification?

I have a multiclass classification problem and decide to use one vs all logistic regression. Since some classes are very rare (pos vs neg is like 1:100), I plan to use some balancing strategy during ...
8
votes
1answer
98 views

Why is the definition of a consistent estimator the way it is? What about alternative definitions of consistency?

Quote from wikipedia: In statistics, a consistent estimator or asymptotically consistent estimator is an estimator—a rule for computing estimates of a parameter $θ^*$—having the property that ...
0
votes
0answers
28 views

Adding weights to data points in logistic regression

I am trying to run a logistic regression on a data sample where the unique identifier is "project". I also have the date on which each project was created. Some projects are more recent than others ...
0
votes
0answers
10 views

Normalisation formula applicable to one or more data items

I've created a recurrent neural network, to which normalised values are passed as inputs. The normalization formula is: $$\tilde{x_{i}} = \frac{1}{1+exp(-\frac{x_i-\bar{x}}{\sigma})},$$ where ...
0
votes
0answers
12 views

Using PyBrain after training a network

I'm using PyBrain to create a neural network. I'm still pretty new to neural networks and their concepts. I've so far only run train() over the network, as ...
1
vote
1answer
15 views

How to get random classification to assess the performance of classifier with McNemar test?

I'm trying to replicate a study where the author used the McNemar test to assess the performance of classification compared to random classification. I have the original classifier and I'm using R to ...
0
votes
0answers
13 views

How to categorize classifiers and matrix factorization methods?

I have a classification problem which is solved by a variety of methods. Among the methods are unsupervised methods, traditional classifiers and a supervised matrix factorization methods. The problem ...
1
vote
1answer
22 views

How do I incorporate the biases in my feed-forward neural network

I'm trying to implement a FFNN. I'm doing this as an excercise to understand how biases play a role in the classification. I trained a NN using a package in R with the inputs being 1..100 and the ...
1
vote
2answers
38 views

Why Adaboost with Decision Trees?

I've been reading a bit on boosting algorithms for classification tasks and Adaboost in particular. I understand that the purpose of Adaboost is to take several "weak learners" and, through a set of ...
1
vote
1answer
66 views

Gaussian is conjugate of Gaussian?

Someone told me that, Gaussian distribution is conjugate to distribution because a Gaussian times a Gaussian would still be Gaussian distribution ? Why is that ? Say the following situation: $X\sim ...
0
votes
0answers
13 views

Predicting the near-future values using an unevenly sampled time-series data

Summary Need help with predicting the near-future values using an unevenly sampled time-series data. Data is collected as events, and is converted to time series. I have tried out a few approached ...
1
vote
0answers
15 views

Why all coeficents of features of model are zero while I have high deviance using glmnet?

I'm using gmlnet to learn lasso regression model. model<-cv.glmnet(x, y, alpha=1, nfolds=10,parallel= TRUE) when I learn model and look at the model it's like this : ...
0
votes
0answers
22 views

Machine learning and Partial differential equations

Are there any algorithms which were developed using partial differential equations for tackling some of the machine learning problems? Most works I see online are in the field of computer vision and a ...
0
votes
0answers
13 views

data amplification

I have obtained few data points from my client, and want to increase the number of data points keeping the complexity of the original almost same. and in the second part increasing the data ...
1
vote
0answers
24 views

A measure of correspondence between ranked ordinal data

I would like to find an appropriate way to measure the similarity between two sets of data with the following characteristics: Both sets contain 10 categorical observations. The categories ...
1
vote
1answer
42 views

How to handle missing data in a small $n$ large $k$ machine learning scenario?

I have a sample size $N=130$ and $1000$ variables. I am using machine learning techniques (SVM) for analysing the data. Some variables in the dataset have values that are so huge that they must be ...
0
votes
0answers
10 views

How process mining compares to probabilistic model (PGM)?

process mining can discovery graph model of process, so can probabilistic-graphical-model (...
0
votes
0answers
22 views

R, Confusion Matrix in percent [migrated]

In R how to get Confusion Matrix in percent (or fraction of 1). The "caret" package provides useful function but shows absolute number of samples. ...
0
votes
0answers
9 views

Standard error of prediction MARS splines earth package

I'm using the earth package (using caret train function) MARS spline implementation in order to perform non - linear regression modeling. I would like to obtain a measure of prediction uncertainty ...
1
vote
1answer
43 views

Ratio between positive and negative examples in a training problem

When training a 0/1 classifier, what should be the ratio of positive to negative, how to decide the ratio between them based on the classifier I use and the data set under analysis?
0
votes
0answers
7 views

Grouping team features and comparing them to a single match outcome

I have two teams, and both teams have multiple features related to that team. For example: Players in the team. Players total points won. Team win percentage. Average player weight. Average player ...
0
votes
0answers
12 views

Is the approach for PLSDA for categorical variables the same as that used for “PLS for regression”?

I understand the approach used for partial least squares for regression (PLS) where the principal components are chosen such that the correlation between the scores in the principal component space ...
0
votes
0answers
31 views

What's an intuitive explanation of how the Otsu's method works? [closed]

i would use otsu's method for reduction of a gray level image to a binary image but i don't understand how it works,i need your help
0
votes
1answer
19 views

Label propagation in semi-supervised learning

Suppose we have a set of labeled and unlabeled instances. 70%unlabeled 30% labeled. We apply a semi-supervised algorithm. Let's say we apply S3VM or Laplacian SVM. We use all the data available. When ...