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

learn more… | top users | synonyms (1)

0
votes
0answers
13 views

What is the derivative of a matrix with respect to f? The matrix has softmax function in it

I need the derivative of W, which has the following expression. W is matrix which contains entries from softmax function. I couldn't find the derivative.
0
votes
1answer
12 views

Grid search error in LIBSVM while optimizing C and g parameters

I am using libsvm for a one-class classification problem. I am trying to select the ideal C and gamma parameters for different kernels(polynomial, linear and rbf) I am using the suggested matlab code ...
0
votes
0answers
16 views

How to standardize text data for training Neural Networks?

I want to train neural network with text data(natural language) as input for classification purpose. One way for standardizing text data for neural network is to use N-GRAM/SKIP-GRAM representation ...
0
votes
2answers
61 views

Instance weighing in libsvm/liblinear

I often use the instance weights with Libsvm for classification problems. http://www.csie.ntu.edu.tw/~cjlin/libsvmtools/#weights_for_data_instances Does anyone know the details of the algorithm that ...
2
votes
1answer
129 views

Machine-learning input data distribution

I'm trying to build a binary 1/0 ML classification algorithm, and was thinking about how to set up the input dataset. If the event I want to predict (the 1's) occur relatively less frequently in the ...
0
votes
0answers
15 views

Identifying topic name

When we perform Latent Dirichlet allocation on a set of documents to cluster, just assume give n=5 topics. How can we assign the topic name for these topics after performing LDA? I would like to give ...
0
votes
0answers
5 views

How to shorten the detection time of adaboost algorithm?

I'm working on a license plate detection using OpenCV's adaboost algorithm. However, after training, it shows that the detection takes 3200ms for a single image, where the image size I used is ...
0
votes
0answers
5 views

Clusteriod questions

I would like to clear some things up because I'm confusing everything. A $clusteriod$ is a coordinate for the mean value of a cluster? So if I have a 2-d .csv file I wish to perform kmeans, the ...
1
vote
2answers
549 views

LibSVM weights of support vectors

I am using LibSVM classifier in my Java code and I am getting correct results as I verified that with weka GUI, however, when I want to get the weights of the ...
0
votes
2answers
71 views

How can I evaluate my model when the the testing data are too few and the generated results don't match testing data

I have a bioinformatic data set includes a very large negative examples (let's say 30000 examples) and just a few positive examples (let's say 150 positive examples). Since I need to feed an enough ...
0
votes
1answer
21 views

A single document as input to LDA?

We use topic modelling usually on a collection of documents - which makes the input. But what if I only have a single document where I want to see the underlying topics in it? I have heard that you ...
0
votes
0answers
8 views

Diference between glm and bigglm estimates

How does bigglm function in biglm package work for logistic regression? I thought that it is not possible to calculate LR on chunks of data and then merge results. Will glm and bigglm yield ...
9
votes
3answers
972 views

Why would anyone use KNN for regression?

From what I understand, we can only build a regression function that lies within the interval of the training data. For example (only one of the panels is necessary): How would I predict into the ...
3
votes
1answer
59 views

“…if the data is linearly separable”

I keep hearing this phrase as a precursor to many algorithms, but I am not sure how exactly one goes about finding out if the data is indeed, linearly separable. Of course, if the data has ...
0
votes
0answers
20 views

ANNs and mixed data-type problems

I did some research but I'm not quite sure if ANNs, more precisely MLPs, are able to handle mixed data-types (e.g ordinal and metric scaled variables) like in the German Credit data set without ...
0
votes
0answers
4 views

EM/K-means task at hand/confusing

Hello I am getting into machine learning and patter recognition, however it's still quite a jungle at the moment. I am using WEKA and Java to try and create my first program. The following is what ...
0
votes
0answers
5 views

XOR backpropagation convergence

I've implemented 3 supervised training algorithms: rprop, online- and batch backprop with momentum. I have the simple XOR test, and I measured how many times they converge out of N iterations. My ...
0
votes
2answers
96 views

Using decision trees to make a binary decision

I have a button that I can press or not press, a binary target that I would like to be 1 as often as possible, and a bunch of features. I also have a bunch of (feature, button choice, target) data, in ...
4
votes
1answer
78 views

Deep learning algorithm

What's the difference between deep belief network and deep convex network?
0
votes
1answer
24 views

Machine Learning approach to solve this selection problem?

I have a set of items S. items can be joined to groups consisting of up to x items. For a group of items i can derive a score Y using some unknown performance measure. What would be the most efficient ...
0
votes
1answer
46 views

Types of artificial intelligence with good results [on hold]

I have been looking into artificial intelligence for some time now. I am wondering what branches are still in active research and have some good/interesting results. The two that I have looked in so ...
3
votes
2answers
136 views

Newbie to neural networks

Just starting to play around with Neural Networks for fun after playing with some basic linear regression. I am an English teacher so don't have a math background and trying to read a book on this ...
1
vote
0answers
7 views

How can recurrent neural networks be used for sequence classification?

RNN can be used for prediction, or sequence to sequence mapping. But how can RNN be used for classification? I mean, we give a whole sequence one label.
0
votes
0answers
7 views

How to Mine Tree Structures?

To learn similarities/differences between different instances (that are in the form of tree), what are the suitable methods/approaches? I know kernel methods and particularly tree kernels, but would ...
0
votes
1answer
333 views
0
votes
0answers
5 views

ImageNet: what does top-five error means?

One of the evaluation method for ImageNet Competition (classify 1,000 categories images) is top-5 error, what does that mean? See: http://www.image-net.org/challenges/LSVRC/
0
votes
1answer
156 views

Machine learning with trinomial features

I have 100,000 students who have each answered some multiple choice questions. Given their performance I want to work out what the chances are of a particular student answering the next question ...
1
vote
2answers
83 views

PCA on train and test datasets: should I run one PCA on train+test or two separate on train and on test? [duplicate]

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 ...
6
votes
1answer
829 views

PCA and the train/test split

I have a dataset for which I have multiple sets of binary labels. For each set of labels, I train a classifier, evaluating it by cross-validation. I want to reduce dimensionality using principal ...
2
votes
1answer
205 views

Comparison of two classifiers based on precision/recall/F1 only?

For two classifiers h1 and h2 I have the precision, recall and F1 score as a percentage (along with the original labeled data set that they were tested on). If I had access to which samples each ...
0
votes
1answer
26 views

Plot that shows which attribute has the most effect on class?

I'm playing around with two datasets: Mushrooms and Breast Cancer. I'm trying to form a hypothesis of which attribute would have the most effect on the class when making predictions about the class. ...
0
votes
1answer
45 views

Gradient of Log-Likelihood

Considering the following functions I'm having a tough time finding the appropriate gradient function for the log-likelihood as defined below: $a_k(x)=\sum_{i=1}^D w_{ki}\cdot x_i$ $P(y_k|x) = ...
0
votes
0answers
5 views

Nonnegative Matrix Factorization coefficient matrix normalization in microarray data

I am confused by the coefficient matrix in nonnegative Matrix Factorization. Suppose I have a time course microarray data matrix as V (m*n), which is decoded by ...
1
vote
1answer
15 views

Distance from hyperplane in SVM rbf kernel in R

I am running ksvm in R(using kernlab package) for a highly imbalanced data.Is there any way i can get the distance of my test data points(each of dimension 8-10) from the hyperplane?so that i can ...
1
vote
1answer
81 views

What are the general strategies in creating a Probabilistic Graphical Model?

While there is lot of theory and probability in the background to understand, I wanted to know if there are any resources/quick pointers on what to consider while modeling a problem using Bayesian ...
1
vote
0answers
16 views

Confidence Interval of Calculating Precision

The problem I'm trying to solve has the following setting: Let $X_1, \dots, X_N$ be a data set where each $X_i \sim Categorical(p_{TP}, p_{FP}, p_{TN}, p_{FN})$ s.t. $\sum\limits_{k \in ...
1
vote
0answers
21 views

How to fit a single quadratic term to a regression

I have a high dimensional multivariate model and am fitting linear weights to each of the $N$ free variables using a classic stable SVD matrix solver. This works. I want to improve the fit by using a ...
0
votes
0answers
10 views

What techniques are used to prevent overfitting in DSN

A Deep Stacked Network (DSN), is a ensemble learner, which roughly works by training a single hidden layer neural network on the inputs and target outputs, then training another which takes an input ...
3
votes
1answer
12 views

Non parametric estimators for noisy functions

Suppose there is a function $f(a,b,c,\ldots)$ of $M$ variables (fixed numbers, not random variables). Add some Gaussian noise to this function: $$ g(a,b,c,\ldots) = f(a,b,c,\ldots) + ...
1
vote
1answer
28 views

How to use geometric proximity in classification

I am doing a classification of certain regions of an image. Let's say I have done the classification, and some classes have been classified positively (negatively) with high probability. For my ...
4
votes
2answers
81 views

Classifying by performing PCA for two classes separately

I have a dataset with binary labels, and I try to figure out whether the data can be classified and yield the ground-truth labels. I thought to try PCA for the data with each of the labels, and see ...
0
votes
0answers
21 views

How can i do time series forecasting with missing data

I am relatively new to time series forecasting, I have worked previously with continuous data at regular intervals successfully, Now I have a data set with missing values, for example look at the ...
0
votes
2answers
584 views

SVM using RBF and nearest neighbor classification method

SVM using RBF kernel is claimed to be similar (equivalent) to the K nearest neighbor classification method. I am not very clear about the analysis process of building this kind of relationship. Thanks ...
48
votes
6answers
32k views

What is the difference between test set and validation set?

I found this confusing when I use the neural network toolbox in Matlab. It divided the raw data set into three parts: training set validation set test set I notice in many training or learning ...
49
votes
8answers
7k views

Skills hard to find in machine learners?

It seems that data mining and machine learning became so popular that now almost every CS student knows about classifiers, clustering, statistical NLP ... etc. So it seems that finding data miners is ...
4
votes
3answers
298 views

What impact does increasing the training data have on the overall system accuracy?

Can someone summarize for me with possible examples, at what situations increasing the training data improves the overall system? When do we detect that adding more training data could possibly ...
2
votes
1answer
65 views

Why doesn't it make sense to preprocess data with PCA before classification?

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 ...
1
vote
1answer
16 views

How do I view “error correct learning” in ANN as an optimal control problem?

There is a lot of material out there for the gradient descent method used in ANN but no body makes it clear how this is an optimization problem or brush it off as extraneously info. Can someone make ...
0
votes
0answers
8 views

Classification problem with constraints

I am trying to solve a classification problem with constraints and need advice on how I should approach it. Here's the problem: Given N observations, FLAG_j, j=1,..,N (this is a binar variable), and ...
15
votes
2answers
721 views

What exactly is a Bayesian model?

Can I call a model wherein Bayes' Theorem is used a "Bayesian model"? I am afraid such a definition might be too broad. So what exactly is a Bayesian model?