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

learn more… | top users | synonyms (1)

0
votes
1answer
38 views

Why do RNN's have a tendency to suffer from vanishing/exploding gradient?

The title says it all: Why do RNN's have a tendency to suffer from vanishing/exploding gradient? For what is a vanishing/exploding gradient see: ...
0
votes
0answers
14 views

When does weight sharing in RNN's make sense (or not)?

It is my understanding that RNN's share weights. It seems to me that this may not be wise for all situations. So if you use an RNN (with weight-sharing) what are you assuming about the problem you are ...
1
vote
0answers
27 views

Bias and variance of a naive bayes classifier and KNN classifier

After reading the paper by J. Friedman, ”On bias, variance, 0/1-loss, and the curse-of-dimensionality,” Data Mining and Knowl- edge Discovery, 1997. I would like to estimate both bias and variance ...
4
votes
0answers
23 views

VC dimension of regression models

In the lecture series Learning from Data, the professor mentions that the VC dimension measures the model complexity on how many points a given model can shatter. So this works perfectly well for ...
0
votes
1answer
14 views

R's nnet needs decay to perform with sin() like function, why variant reproducibility

I've noticed that for sin() like data, I need to use "decay" which is available in nnet to get the ANN to perform. Why would that be in theory? Also when I run runNN(0.02) over and over, sometimes ...
0
votes
1answer
22 views

Question about Continuous Bag of Words

I'm having trouble understanding this sentence: ...
3
votes
1answer
61 views

How do you transform a decision boundary in the angle kernel to the original space?

Say I have training data $S_n$ and each point is of the form $x = \langle x_1 , x_2 \rangle$ in the original space (i.e. $x^{(i)} \in \mathbb{R}^2$). I was considering the following kernel: $$ ...
2
votes
3answers
98 views

Machine learning for discovering formulas

I have a vector of desired values that I want to fit based in some generated predictors. The tricky part is that I wish to have the explicit formula. For example, giving the below input I would like ...
1
vote
0answers
22 views

Python kNN vs. radius nearest neighbor regression

Python offers two nearest neighbor regressions: radius nearest neighbor and k-nearest neighbor. I'm trying to figure out a few things: 1. Under which circumstances would each be preferable? 2. How do ...
0
votes
0answers
20 views

Can I use the ECDF of a ~NB variable for SVM?

I am hoping to use SVM or NN for a large dataset (N ~ 4-8 million). Some of the features are counts, which have a distribution that is approximated by the Negative Binomial in the bioinformatics ...
0
votes
1answer
30 views

Model comparison between glm (with Firth correction), random Forest, penalised SVM

I am currently developing three models to classify features of gene sites. I was using glm (with Firth correction), random Forest and SVM to build the models and I used forward and backward ...
1
vote
0answers
91 views

Statistical modeling method for curves with uncertainty

I would like to ask for advice on choosing a suitable modelling method for the following problem: I am modeling the performance of a device for curve estimation. I have collected a data set ...
0
votes
0answers
19 views

Sensitivity analysis of machine learning techniques

As you know we can have sensitivity analysis (sensitivity of output(s) based on changing of inputs) in different kinds of regression. Can we have sensitivity analysis for machine learning techniques ...
0
votes
0answers
8 views

classification when some classes can be grouped

Is there a systematic way to address a classification problem where some classes are dependent and thus can be clustered to construct a larger superclass? I gave an example in this Classification ...
1
vote
0answers
27 views

What is best practise for dimensionality reduction in rows of data

I was wondering what was best practise for dimensionality reduction in observations (as opposed to features) in a data-set? I often have data comprising of a multiple, random number of observations ...
0
votes
0answers
21 views

why boosting method is sensitive to outliers

I found many articles showing that boosting methods are sensitive to outliers, but no article explains why. In my experience, I feel outliers data is bad for any machine learning algorithms, but why ...
0
votes
0answers
29 views

Hidden Markov Model For Text Classification

I have a question about HMMs being used to classify an entire text body under examination. This is as opposed to classifying a subset of a text body under examination. For example, classifying a news ...
2
votes
1answer
38 views

Classification when some classes are dependent

I think my problem can easier be explained via an example: Assume we have a dataset containing the images of 10 different mammals, let's say lion, elephant, cat, ... and horse. We have a 20-class ...
0
votes
0answers
10 views

What is an “example based” metric?

In the context of multilabel learning I came across several "example based" metrics, for instance example based Recall, example based Precision etc. (see here) I do know the concept of Recall etc. ...
2
votes
0answers
58 views

preferential multinomial model (with memory)

I am modeling a system that is like having a container of an infinite number of colored balls. On each day $t$, I pull out a new set of $n_t$ balls and I count the number of balls that are red, green, ...
0
votes
0answers
33 views

using biomod2 package with continuous response variables

I am looking to correlate crop area with climate variables and then predict for future if suitability of crop areas will change/remain same under different climate scenarios (in line with species ...
0
votes
0answers
12 views

evaluating the performance of item-based collaborative filtering for binary (yes/no) product recommendations

I'm attempting to write some code for item based collaborative filtering for product recommendations. The input has buyers as rows and products as columns, with a simple 0/1 flag to indicate whether ...
1
vote
0answers
34 views

What are the latest developments in new statistical learning algortithms? [closed]

I was trying to implement statistical learning algorithms in my research. I started with Artificial neural network but later I found out that Boosted decision trees was much better for the task. Now ...
1
vote
1answer
39 views

How to do machine learning (regression/classification) when the samples are of different sizes?

In standard cookbook machine learning, we operate on a rectangular matrix; that is, all of our data points have the same number of features. How do we cope with situations in which all of our data ...
0
votes
0answers
32 views

RandomForest Classifier With Very High Success Rate

I'm having a weird problem that may suprise you all. My classification rate is too high on my test set. I'm using scikit-learn packages, and I'm very suspicious of these classification rates, as they ...
2
votes
1answer
74 views

How to score predictions in test set taking into account the full predictive posterior distribution?

I have three predictive models (regressions) which parameters are estimated by Markov Chain Monte Carlo. Predictions are made over a test set of size $N$. Since I compare the models under different ...
1
vote
0answers
33 views

When to use Bayesian Networks over other machine learning approaches?

I expect there may be no definitive answer to this question. But I have used a number of machine learning algorithms in the past and am trying to learn about Bayesian Networks. I would like to ...
0
votes
1answer
38 views

Can we boosting or stacking with different input variables for each model in machine learning?

I have a question about Boosting and stacking in machine learning. Suppose that I will train neural network, SVM and logistic regression using optimization algorithm to optimize best inputs in first ...
0
votes
0answers
11 views

Gaussian Mixture Model with Custom Distance Metric

I have some 1D data that I want to cluster using Mixture of Gaussian. However, the data "wraps around" at two extremes. Specifically, I have a list of angles from $-\pi$ to $\pi$ and the data near two ...
1
vote
1answer
143 views

Why is there a E in the name EM algorithm?

I understand where the E step happens in the algorithm (as explicated in the math section below). In my mind, the key ingenuity of the algorithm is the use of the Jensen's inequality to create a lower ...
0
votes
0answers
11 views

finding weights of aspects

I have aspects(nouns) from online customer reviews of a product. I have done sentimental analysis to get polarity of each aspect. Now I want to weight the ...
3
votes
1answer
89 views

When should I use feature selection and when should I use dimensionality reduction techniques?

When should I use feature selection and dimensionality reduction? I know that feature selection is different from dimensionality reduction. But I don't know under what circumstances should I use ...
0
votes
0answers
12 views

How to apply AIC to a situation where the mean of a multivariate normal is a 0-1 d-dimensional vector with exactly k 1's

I am trying to apply AIC to estimate mean in the following case: Let us consider that I have $n$ random variables $X_1, \ldots, X_n$, drawn i.i.d. from a normal distribution of mean $\mu\in\{0,1\}^d$ ...
0
votes
0answers
4 views

Weka - StringtoVector Filter Not working [migrated]

I am practicing Weka using the Reuters data. The StringtoVector Classifier works for converting my string data (shown below), so I can analyze the articles to understand what words predict the ...
0
votes
0answers
10 views

Comparing the impact of 2 independent variables on the dependent

I'm using a predictive modelling technique which has 2 parameters. I've performed a sweep of values for each of these 2 parameters, running each permutation of parameters 30 times as the technique is ...
1
vote
1answer
35 views

Regressing a discrete variable

I have a discrete dependent variable (say, number of units bought) and want to run a linear regression with in-store promotion, seasonality, trend etc. as predictor variables. I'm not sure if it is ...
1
vote
1answer
43 views

basic question about machine learning and probabilistic framework?

Why do we assume that the pair (X,Y) where X and Y are features and labels respectively are random variables governed by a probability distribution? Why does this assumption make sense? What if the ...
4
votes
0answers
66 views

Why would a regression model predict super huge numbers? [closed]

I have a set of 55 items. Each item is defined by 6 values. I am doing 55-fold cross validation: training a model on 54 items, predicting on the 55th. The 6 values of the 54 items are used in some ...
0
votes
0answers
19 views

What are the advantages of using logistic regression with kernel over others?

What are the advantages of using logistic regression with kernel over others type of logistic regression(e.g.,dot)?
1
vote
2answers
90 views

How to transform categorical variable into numerical variable when using SVM or Neural Network

To use SVM or Neural Network it needs to transform categorical variables into numeric variables, the normal method in this case is to use 0-1 binary values with the k-th categorical value transformed ...
1
vote
0answers
16 views

Item based collaborative filtering when items have been available for different lengths of time

I am attempting to use item based collaborative filtering for product recommendation. The matrix is all 1s and 0s based on whether or not a buyer purchased an item, and I am using cosine similarity to ...
0
votes
0answers
5 views

How to handle multiple experiments or same row/entity?

I have data that was gained from experiments preformed by n individual persons. So for each item in the experiment I have n values for the same variable. Note that IMHO averaging the values does not ...
0
votes
0answers
27 views

what's the best empirical macro/micro F1 score?

Theoretically it should be 1. In the following presentation it's said that "0.5 to 0.55 (micro) F1 score is best for multilabel classification problems" I tried to investigate this statement but ...
0
votes
0answers
23 views

Are there differences between Delta TF-IDF and TF-IDF?

Are there differences between both algorithm or not ? i mean if i implement for ex Delta TF-IDF in a project instead of ...
3
votes
1answer
76 views

Seeking a free symbolic regression software [closed]

Now that Formulize / Eureqa started charging $2500 a year for using it and having crippled the trial version, does anyone know of any replacements that can do similar things like find an equation ...
0
votes
0answers
19 views

Perceptron Algorithm, RMSE just cycles through two numbers

My input is a bag of words feature vector, of the form: Example: Document 1 = ["I", "am", "awesome"] Document 2 = ["I", "am", "great", "great"] Dictionary is: ...
1
vote
0answers
27 views

Testing hypothesis when each data instance is a vector

The experiment involved studying population body velocity changes during a marathon . Velocity of each person in control group and experiment group were measured at multiple instances during marathon. ...
2
votes
1answer
25 views

How to handle changing input vector length with neural networks

I want to train a neural network with a sequence of character as an input vector. Learning examples have different length and for this reason I don't know how to represent them. Let's say I have two ...
0
votes
1answer
39 views

What is the difference between independent variable and a feature?

I ran into this question which asks the identification of various terms for a linear regression function (f). I am confused about the "independent variable" definition. What is the difference ...
0
votes
1answer
79 views

How to pass the sequential var. length data for the NN?

The main task is from Inductive Logic Programming (ILP) area. The task related to ANN is inspired by paper below but is applied to more complex case.Learning an approximation to clause evaluation ...