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

learn more… | top users | synonyms (1)

0
votes
0answers
11 views

Changing the class file from nominal to numeric in order to carry out the linear regression [closed]

I am trying to perform the LinearRegression classification on the diabetes.arff and glass.arff data sets, with little luck. The option is greyed out and I cannot ...
8
votes
5answers
299 views

What algorithm should I use to detect anomalies on time-series?

Background I'm working in Network Operations Center, we monitor computer systems and their performance. One of the key metrics to monitor is a number of visitors\customers currently connected to our ...
3
votes
0answers
21 views

Bound on the total change using Pearson's r

I am given an increasing series $(x_1,....x_n)$ and I know the pearson correlation between $(x_1,....x_n)$ and some (unknown) increasing series $(y_1,....y_n)$. Can I derive an upper and a lower ...
0
votes
0answers
13 views

Interpret F values of selected features

I have a dataset that contains wines and their ratings. An entry contains the name of the wine, the grapes used, the year and the rating: 'Chateau Pape', 'Pinot Gris', '1983', 93.4 I'm interested in ...
2
votes
1answer
84 views

What should be my training set for a word-spotting program?

I wanted to create a simple voice recognition program. It should be able to distinguish my voice from other sounds. Aside from my voice, what should I add to my training set? Also, I want to extend my ...
-1
votes
0answers
34 views

How to report machine learning research?

I am using support vector machines and cross-validation for a binary classification task. I have constructed three different models, and therefore I have three sets of results. How should I report the ...
0
votes
0answers
23 views

Feature selection of non stationary data

I am working with EEG signals which are non-stationary. I have used spectrogram to analyse the data in specific frequecies. I have to select some features from the specific-frequency time signals ...
0
votes
0answers
14 views

Longer forecasting with one-step-ahead model

It is totally a noob question but I cannot find any explanation on the subject. Suppose I build a forecasting system for time series $x$, using as inputs $[x_{t-n},...,x_t]$ to predict the next ...
0
votes
2answers
42 views

Will decision trees perform splitting of nodes by converting categorical values to numerical in practice?

In Decision trees, while doing classification or regression, are we using only numerical values. Suppose if i am having a column of 'Wind' as a feature. Suppose, I am having 5 rows ( observations ). ...
0
votes
0answers
17 views

Minimum training sample size for a hierarchical hidden semi-Markov Model (HHSMM) [closed]

I've implemented a HHSMM machinery but I have doubts about the minimum size of training set that I have to acquire for the experiment. Any method?
4
votes
1answer
36 views

Does rank of observation matrix tell anything useful when applying machine learning?

Suppose I have an observation matrix of size $N \times M$ where $N$ is the number of samples and $M$ is the number of variables. If the rank of the observation matrix is $R<M$, does it tell ...
0
votes
1answer
8 views

GPML producing wrong output using correct target labels

I am using the GPML code found here. The key function in the aforementioned library is the gp function described below: Two modes are possible: training or ...
1
vote
0answers
67 views

Error in linear regression

Given two series $(x_1,...x_n)$ and $(y_1,...y_n)$, and assume that we know $x_{n+1}$. Given the fact that the pearson correlation won't change in the next observation of $y_{n+1}$, can we bound the ...
1
vote
0answers
20 views

Difference between Latent and Explicit Semantic Analysis

I'm trying to analyse the paper ''Computing Semantic Relatedness using Wikipedia-based Explicit Semantic Analysis''. One component of the system described therein that I'm currently grappling with ...
0
votes
0answers
37 views

Forecasting values based on day of week and hour

Disclaimer: Not really good at statistics. Scenario: I have some data, sampled by the hour. I want to make some forecasts taking into consideration that the data seems to be influenced by the day of ...
0
votes
0answers
8 views

UCI Machine Learning Data Set: Parkinson Speech Dataset with Multiple Types of Sound Recordings Data Set

I would like to use the data set Parkinson Speech Dataset with Multiple Types of Sound Recordings Data Set from UCI to test pattern recognition algorithms. However when I plot the features and ...
0
votes
0answers
20 views

learing structure of Bayesian Network from data

If I have data and I want to learn the structure of Bayesian Network from these data but the node ordering is not given which algorithm I should use.
0
votes
0answers
13 views

Pairwise Learning to Rank - detecting detrimental changes

The idea behind Pairwise Learning to Rank is that if you have a set of search results then a clicked on result can be used as training example to indicate that it should rank more highly that the ...
2
votes
0answers
24 views

What does improper learning mean in the context of statistical learning theory and machine learning?

I was reading the following paper and it talked about improper learning. I wasn't 100% what it rigorously meant but they do mention: I am not sure what "representation independent" means, but as ...
0
votes
1answer
57 views

Using Neural Networks to predict stock values

How are neural networks usually used to predict market evolution? My data consists of a set of pairs (time, value), taken at an interval of 15 minutes. My ideas so far are: I.Take 40 values (or ...
7
votes
2answers
120 views

How to make sure that a machine learning algorithm's implementation is correct?

Say there is a machine learning algorithm (e.g. classification) that is well known and implemented by the original creators of the algorithm. Yet all you have is the ability to use the algorithm but ...
0
votes
2answers
55 views

Data splitting and cross validation

my question is about splitting data! I used to split data into training and testing set using caret library in R ...
0
votes
0answers
24 views

Caret feature selection [RFE] yields different features depending on reference level of binary outcome

I'm using RFE from the caret package in R to select variables to be used in a linear discriminant analysis. The outcome is a binary factor, but depending on which level of the factor is used as the ...
1
vote
1answer
33 views

distinction between Bayesian Network and another graph [closed]

I want to ask if we have two graph one of them Bayesian Network and the other one just regular graph, how we can distinction between them.
0
votes
0answers
13 views

Difference in training procedure for DBN and DBM

This is related to the following thread Deep belief networks or Deep Boltzmann Machines? but it doesn't seem to answer in a practical sense what the difference is. So I gather a DBN is directed and ...
5
votes
1answer
96 views

Does the opposite of nested cross-validation make sense?

I'm asking the question from a machine learning point of view. I have a dataset with relatively high sparsity, so if I use nested cross-validation for my feature tuning and evaluation, that is tune ...
3
votes
2answers
282 views

My Test accuracy is pretty bad compared to cross-validation accuracy

I did a Multi-class document classification. I divided the original data set (18,8334 documents as a list of strings where each element of list is a document string.) into 70% training and 30% test. ...
2
votes
3answers
64 views

Criteria for classification performance

In binary classification, are there criteria or guidelines available to judge if classification performance of the testset (unseen data) is poor, medium or high? I realise that this may depend on ...
3
votes
1answer
70 views

Machine learning for pattern recognition in realtime sensor data

I'm working on a project where we need to detect patterns in a sensor's output to find out if a given event occurred. Given my limited experience with machine learning, I was wondering if someone ...
0
votes
0answers
25 views

customer analysis - book / blog recommendation?

I'm new to the topic 'customer analysis' (in general) and need advice for a good starting point: What is a good book / blog / tutorial on this topic? My current situation is: I have a lot of customer ...
1
vote
1answer
22 views

Class-specific feature importance

I have rather a simple question which I have not had any luck finding the answer to. I'm training a Random Forest classifier using sklearn in Python 2.7, on a large dataset ~(80k,250) where ...
0
votes
0answers
13 views

MANOVA application

Need to check whether it is mandatory to have at least one continuous DVs in MANOVA, also does sample size of 2 groups matters much, i am planning to check whether recruits from campus or off-campus ...
1
vote
2answers
49 views

Nested cross-validation - how is it different from model selection via kfold CV on the training set?

I often see people talking about 5x2 cross-validation as a special case of nested cross validation. I assume the first number (here: 5) refers to the number of folds in the inner loop and the second ...
2
votes
2answers
59 views

Why don't people use deeper RBFs or RBF in combination with MLP?

So when looking at Radial Basis Function Neural Networks, I've noticed that people only ever recommend the usage of 1 hidden layer, whereas with multilayer perceptron neural networks more layers is ...
0
votes
0answers
17 views

Algorithms for Suggesting Users

There are often a large number of bug tickets that get generated for the projects at my company. I was thinking of an intelligent way for automatically suggesting users to assign to the tickets based ...
0
votes
0answers
23 views

NaiveBayes, J48 and RandomTree in layman's terms

I am difficulty understanding how both classifiers work under the hood. So far I have deduced NaiveBayes predicts an outcome by 'uncoupling' multiple pieces of evidence, and to treating each of piece ...
0
votes
0answers
20 views

Huffman tree generation if the frequency is same for all words [migrated]

Can a valid Huffman tree be generated if the frequency of words is same for all of them? Example : ...
0
votes
0answers
13 views

Implementing PCA using Incremental approach [migrated]

I am trying to implement the algorithm proposed in the paper in Section (III) here in R. It uses incremental eigendecomposition and incremental SVD for calculating IPCA. Instead of working on images ...
1
vote
2answers
42 views

Knn classifier for Online learning

Is Knn classifier suitable for online learning i.e. Is it effective to apply online learning approach for knn classifier?
0
votes
1answer
35 views

Backward propagation algorithm demonstration in neural networks: any VERY-SMALL-STEP by VERY-SMALL-STEP demonstration?

I'm looking for a VERY DETAILED demonstration for the backward propagation algorithm in neural networks machine learning. Specifically the step below. I've got the excellent Michael Nielsen ...
1
vote
0answers
24 views

When is logistic regression minimizing under squared error loss the same as maximizing binomial likelihood?

Implementing logistic regression and getting different results depending on whether I minimize squared error or maximize log likelihood. When are the two equivalent?
3
votes
2answers
81 views

Are N-1 distinct points with two classes Linearly Separable in an N Dimensional Space

I have this conjecture whose verification I haven't been able to find online anywhere. If you have N - 1 distinct points that each have one of two classes, can you find a linear decision boundary in ...
0
votes
1answer
71 views

Why am I getting 100% accuracy for SVM and Decision Tree (scikit)

I have a dataset with 1175 examples and 21 features which are in the range of [-1, +1], and two class labels 1 and 0. As I read in the most of the resources, it is good to have data in the range of ...
1
vote
1answer
35 views

Choosing right range for data while using scikit-learn

I have a dataset with 1175 examples and 21 features which are in the range of [-1, +1], and two class labels 1 and 0. As I read in the most of the resources, it is good to have data in the range of ...
0
votes
2answers
41 views

Using k-means for reducing the size of the training set of a Kernel SVM

I have a classification problem with the following characteristics: a few million data points around one hundred features non-linearly separable Training a SVM with an RBF Kernel is not feasible ...
2
votes
3answers
49 views

Question regarding form of a cost function while training a model

I am beginner in Machine Learning and I am very interested in modelling, simulation and all this jazz :) One of the basic idea I learned so far is the use Cost Function and its minimization in order ...
0
votes
0answers
42 views

How to measure error in linear regression?

Suppose I have $X, Y =$ {$(x^{(1)}, y^{(1)}); (x^{(2)}, y^{(2)}); ...; (x^{(m)}, y^{(m)})$} examples with: $x^{(i)} \in \Re^n$: for $i$ $\in$ [1;m], $x^{(i)}$ = {$x_1^{(i)}, x_2^{(i)}, ..., ...
-1
votes
2answers
76 views

Large? Number of parameters in MCMC model [closed]

I am implementing a Hierarchical Bayesian Modeling in order to model the relation between the independent and dependent parameters $(x, y)$. I assume the relation is: $$ y_i = \alpha + \beta x_i + ...
2
votes
1answer
246 views

How does the mean function work for a Gaussian Process?

I was reading the notes on Gaussian Processes by Choung B. Do (stanford course CS229) however was unsure of how the mean function worked and what a random variable was on the Gaussian Process So ...
0
votes
0answers
56 views

How to combine weak classfiers to get a strong one?

Let as assume that we have a binary classification problem. We also have several classifiers. Instead of assigning a vector to a class (0 or 1) each classifier returns a probability that a given ...