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

learn more… | top users | synonyms (1)

0
votes
0answers
17 views

help in coding decision tree in python [migrated]

I am not sure if this is the right place to post this, but I have been trying to code up a simple decision tree class for a while and am getting lost at various points. Specifically, I'm not sure ...
1
vote
1answer
63 views

Understanding k-means unsupervised learning for features

I'm following this paper: http://ai.stanford.edu/~ang/papers/icdar01-TextRecognitionUnsupervisedFeatureLearning.pdf And I'm trying to understand specifically how the k-means approach works when ...
1
vote
2answers
57 views

How can I force my model to predict the samples that are close to zero?

I have a large amount of inventory data and I am trying to predict when the inventory gets low using one component of the change in inventory (yes I know this doesn't describe inventory very well by ...
0
votes
0answers
124 views

HMMs with feature vectors (block HMMs?)

I'm quite new to HMMs, but still couldn't find an approach to fit HMMs in R, where we have feature vectors instead of single values. Or perhaps I simply didn't understand some of the proposed ...
1
vote
1answer
54 views

Will Linear Regression choose as good of a model as any other regression algorithm given enough data?

I am in the seemingly unusual situation of having practically unlimited data. In this case, will linear regression choose as good of a model as any other algorithm in the case where the number of ...
0
votes
1answer
49 views

Understanding decision trees

Im new to decision trees and am trying some decision tree modelling now with titanic data. I have the following dataset. ...
0
votes
1answer
34 views

Model based approaches to content based recommenders. How does this work?

I have a question regarding the use of model based approaches to recommender systems. So, the goal is to create a model that predicts the user reaction to a specific item. Either a rating scale or ...
0
votes
1answer
28 views

Methods to reduce the number of class labels

I have a dataset with 17 classes which will be predicted using a classification algorithm (QDA, Decision Tree or something similar) using between 2-4 features. Many of these classes have significant ...
0
votes
0answers
58 views

How to deal with imbalance data in , for example, neural network

Do we usually discard this issue, just train the neural network then compute the AUC, or can we use weighted version of loss function, for example, in binary classification problem, can we use ...
1
vote
1answer
53 views

How to optimize the number of topics using R Mallet

I would like to select the optimal number of LDA topics using R's mallet library. I know that there are several ways to do this using other implementations of LDA ...
0
votes
1answer
28 views

What statistical test should I run to select “explicative” features in my dataset?

I have a database with more than 500 samples with 22 quantitative features each and I would like to predict a categorical variable (0 or 1). I am trying to fit a logistic regression model and a neural ...
0
votes
1answer
26 views

Factored sparse coding: how to differentiate coordinate transform?

I'm trying to implement a sparse coding technique described by Bergstra et al in this paper. The encoding process there includes the following computation: $translate(rotate(scale(\mathbf{u}, \alpha, ...
0
votes
2answers
91 views

From where I can learn to use scikit learn

I have read text about machine learning and I feel that I have gained sufficient knowledge that I can start applying them practically. I have programming experience in python so I want to learn how to ...
0
votes
0answers
36 views

Choosing contrast coding in R

I am working on a data set with categorical variables. To apply ANN, I want to apply contrast coding to those variables. But how do I choose between coding functions in R (sum, helmert, treatment ...
0
votes
0answers
18 views

Statistical/ML models when observations have different amounts of input

Let's say we're predicting an employee's performance review score for the following year based on his/her performance review scores from each previous year of their employment. We might have these ...
0
votes
0answers
15 views

pattern recognition in spatial data in R

I am currently working on telemetry data for the wild animals. I have extracted their paths over the map. I have also calculated the angular movement per activity. In my dataset I am given with the ...
-1
votes
0answers
20 views

K means algorithm on an image

Hi I want apply my k means algorithm on an image to a certain number of k clusters.I want to ask is that How is the algorithm applied ? I mean what i have learned from internet is that k means is ...
0
votes
0answers
7 views

MLP with hidden layer units more than input layer

What is an acceptable scenario in multi-layer perceptron when the number of neurons in the hidden layer is more than input layer neurons. What changes to the cost function should be made for such a ...
2
votes
1answer
21 views

A priori parameters to Random Forest based on n and p

If cross validation is too costly to determine the number of trees and number of max_features is there any standard to what you choose based on n and p. I know sqrt(p) is standard for max_features but ...
0
votes
2answers
70 views

High accuracy during cross validation, low accuracy on test set

I'm currently trying to build a tennis prediction model. Unfortunately, I have some issues that I hope you could help me to handle. I have 1110 examples of matches from the year 2013, with their ...
3
votes
3answers
7 views

How to build a predictive model that uses only a subset of training factors , for testing?

Generally, all the predictive algorithms work as follows : input factors : x1,x2,x3,x4...xn<br> responses: y1,y2..ym <br> and the model (say M) built ...
1
vote
1answer
44 views

PCA to reduce dimensionality then classifier of choice?

I read about PCA online and the way it computes a covariance matrix, computes eigenvalues, and then transforms the matrix to reduce the dimensionality of the matrix to a certain number k <= p. ...
0
votes
0answers
17 views

Incorporating letter transition model into linear-chain CRF

Suppose I have a linear-chain CRF for e.g. handwriting recognition, $$ p(\mathbf{y}\mid\mathbf{X}) = \frac{1}{Z_\mathbf{X}}\exp\left(\sum_{j=1}^m\mathbf{w}_{y_j}^T\mathbf{x}_j + ...
0
votes
1answer
47 views

K-fold repeated cross validation for classification accuracy in Caret

I am new to cross-validation and I have a data-set called LDA.scores for 12 measured call-type parameters. I am trying to run a k-fold repeated cross validation with 10 folds and associated naive ...
0
votes
0answers
21 views

Stochastic Gradient Descent vs Online Gradient Descent

I was wondering what the difference between stochastic gradient descent and online gradient descent is? Or is it the same algorithm? Thanks.
0
votes
2answers
35 views

receiver operating characteristic (ROC) on a test set

The following image definitely makes sense to me. Say you have a few trained binary classifiers A, B (B not much better than random guessing etc. ...) and a test set composed of n test samples to ...
0
votes
1answer
30 views

Imputation and linear regression analysis paradox

Missing values, especially in small datasets, can introduce biases into your model. There are several data imputation methods (MICE, Amelia II), which use EM algorithms to "fill in" the missing ...
1
vote
0answers
21 views

How effective is randomly/algorithmically generating exorbitant amounts of training data for a neural network?

There are some problems of which the generating of data is easy whereas the inverse is not. For example, use a 3D game engine to render some randomly generated objects with some random changes and ...
1
vote
0answers
16 views

10-fold cross validation for forecasting time series with explanatory data ?

I saw that the question was asked some years ago here, but I wasn't satisfied with the answers so I'm asking it again. Is there some theoretical foundations about not doing k-fold cross validation ...
0
votes
0answers
20 views

What is correct implementation of LDA (Linear Discriminant Analysis)? [migrated]

I found that the result of LDA in OpenCV is different from other libraries. For example, the input data was ...
1
vote
1answer
23 views

How to use external data?

I'm building the model using the internal data to predict the health situation of the customer. I've just found about 100 new "external data" in the form of region data. Based on this new information, ...
0
votes
0answers
14 views

On which kind of problems preprocessing data using RBMs (unsupervised) could give an edge?

I am new to machine learning and basically so far I've been using only supervised algorithms, however, recently I started to use RBMs in combination with some classifier (using a pipeline in scikit ...
1
vote
2answers
44 views

In back propagation for neural networks, what exactly is the “error signal”?

For example: Imagine we end up with a sum of 0.755 on our output node. We then apply the activation function (in this case I'll use a sigmoid) which gives us a ...
0
votes
0answers
21 views

What value of (adjusted) McFadden R square or other pesudo R square means good fitting

I got adjusted MaFadden R square for logistic regression: 0.918772 , 0.6135568 , 0.3407252 respectively, which value is good? I just heard the value between 0.2 and 0.4 is good for McFadden R square. ...
0
votes
0answers
14 views

Which model should I use for extracting binary features?

I create some data, a normal matrix: 1, 11, 0, 20 ..... 1 1, 0, 0, 11..... 1 1, 5, 0, 22..... 0 each row include more than 2000 integer. they are following some patterns. The requirement is ...
0
votes
1answer
32 views

What is the difference between data normalization and feature extraction in deep learning?

I am learning about the multi-modal deep learning models and the papers I am reading are very confusing on one point in the process: "feature extraction" and "data normalization" seem to be used ...
3
votes
0answers
60 views

How to build “supervised clustering” for neural networks?

I'm confused as to what the output would be. Consider the "blind source separation" problem. Let's say I have a ton of training examples where the input is the final cacophony of sounds as a sound ...
1
vote
1answer
21 views

Is multiple stage binary classification a good idea if you have very few positives?

The problem is the following: We have a set of, say 5000 documents, with a single binary label. Say that 4900 documents are negative and only 100 are positive. I built a binary classifier while ...
1
vote
0answers
24 views

Any idea about application of deep dream? [migrated]

Recently Google publicized interesting deep dream. Besides art generation such as http://deepdreamgenerator.com/, do you see any potential applications of deep dream in computer vision or machine ...
0
votes
0answers
22 views

How to handle many-to-one data?

I'm trying a kaggle contest, just to improve my machine learning skills. The challenge I currently do involves many-to-one relational data. For instance, a person belongs to a municipality. A ...
1
vote
2answers
55 views

sklearn - overfitting problem

I'm looking for recommendations as to the best way forward for my current machine learning problem The outline of the problem and what I've done is as follows: I have 900+ trials of EEG data, where ...
2
votes
1answer
35 views

Kernel PCA for feature selection for various machine learning algorithms [duplicate]

I would like to forecast stock index returns with SVM, k-NN, and Neural Networks. In advance I want to select my inputs via kernel PCA (kPCA). Everything is performed in R. For the KPCA I use ...
0
votes
1answer
31 views

MNIST dataset black or white background

In the MNIST dataset, are the images on white or black background? I seem to have encounter both type of images by googling around. Does the background color has any effects on the performance of a ...
2
votes
1answer
40 views

How to apply Cross Entropy on Rectified Linear Units

I am currently getting started with Machine Learning. However, I have some problem to derive formula and not able understand how to applied the Cross Entropy (CE) on Rectified Linear Units (ReLU). I ...
1
vote
0answers
29 views

Prediction Interval & Machine Learning

For linear regression, one can calculate a confidence interval for a prediction. Is this possible for all machine learning techniques? For example let's say I fit a neural network, and then start ...
0
votes
0answers
29 views

Neural Net in high dimensions for images

I'm trying to build a neural net for a image recognition problem. My images are way too large to build a straight up NN from just the pixels; they are about (1000, 1000) width,height. So naturally i'm ...
4
votes
3answers
202 views

What happens when I use gradient descent over a zero slope?

Let us assume my cost function such that its slope rises from point A at z=2 to B at z=4; stays constant till C at z=4; falls to D at z=1; rises to E at z=5. If I choose my starting point between ...
3
votes
1answer
75 views

CART: Selection of best predictor for splitting when gains in impurity decrease are equal?

My question deals with Classification trees. Consider the following example from the Iris data set: I want to manually select the best predictor for the first split. According to the CART ...
1
vote
1answer
18 views

Cost Function vs Precision/Recall

This question was also asked on Andrew's Ng Coursera class forum. I just finished watching this video on skewed classes. Andrew talks about an error function, which is not the same as the cost ...
0
votes
0answers
13 views

How to match two fuzzy membership functions with max-min?

I have two triangular membership functions, let's say A=[0.2,0.5,07] and B=[0.4,0.6,0.8] with the same universe X=[0:0.01:1]. Can someone show me numerically how to calculate the matching function ...