Questions tagged [classification]

Statistical classification is the problem of identifying the sub-population to which new observations belong, where the identity of the sub-population is unknown, on the basis of a training set of data containing observations whose sub-population is known. Therefore these classifications will show a variable behavior which can be studied by statistics.

2,260 questions with no upvoted or accepted answers
Filter by
Sorted by
Tagged with
0
votes
1answer
36 views

How can one classify data if the range of possible features is continuous/infinite?

Imagine my instrument encodes colors using any/all possible wavelengths of light (not just the 3 features red, green, blue). Thus, it has an “infinite” number of features lying on a continuous ...
0
votes
1answer
226 views

random forest classification appears dependent on Dependent variable proportions

I'm using Random Forest for classification which gives the following confusion matrix. 0 1 class.error 0 839 24 0.027 1 60 86 0.410 You can notice that the ...
0
votes
1answer
212 views

Classification with partially labelled data - potential positives

I am having trouble figuring out the best approach for a classification problem: My data: For each physician in my data, I have a feature set of every different medical procedure where the feature ...
0
votes
1answer
32 views

Classifying 3 groups statistically on a continuous scale

I have 3 subsets, say- a,b,c which are derived from a single population of subjects. All subsets have an unequal "n", and are accompanied by a variable say 'L'. My aim is to define a metric (based on ...
0
votes
1answer
196 views

Classification: training sets different sizes

I'm building a classifier for text analysis sentiment. I have a large training set for positive, neutral and negative mentions. Should the training data sets be similar in size? Currently my ...
0
votes
1answer
23 views

Classification method with (potentially) endless training input

What is the best multi class classification method with (potentially) endless training input? The classificator should get trained while a user interacts with the system. At this time it gets ~ 30 ...
0
votes
1answer
55 views

How to embed prior distribution of datasets into the classification model

Given 3 training sets : $(X_1,y_1),(X_2,y_2)$ and $(X_3,y_3)$. These three datasets are separated as it is being manually tagged in the preprocessing. Based on the datasets, three classifiers can be ...
0
votes
1answer
61 views

Predicting continuous position using input variables of unknown quality

The problem I'd like to solve can be reformulated as follows. Let's consider that I have to go to some parties and I would like to find out where in the room I am most likely to have a good time. I ...
0
votes
1answer
105 views

How do I learn a simple cut-off value between 2 classes given one-dimensional data?

Given a set of data which consists of a single real number and a class, I want to find a value (i.e., the inflection point if we were talking about logistic regression) which would lie right at the ...
0
votes
1answer
159 views

Conceptual diffiulty in understanding multiclass classification

Given a high dimensional feature vector x in $R^D$, I want to map it to an L bit vector, $L << D$, z = h(x) in $\{0,1\}^L $ using a function h while preserving the neighbors of x in the binary ...
0
votes
1answer
24 views

Classification when variables are observed as a group

How do I classify variables when the classifying binary output is known only for groups of variables? Here is a concrete example: a person eats different types of foods on different days, and she ...
0
votes
0answers
207 views

Does Gaussian discriminant analysis and linear discriminant analysis refer the same algorithm?

I'm pretty new to LDA and I came across other terminology called Gaussian discriminant analysis elsewhere. Since LDA assumes the normality or normal distribution of the data which is same as ...
0
votes
0answers
83 views

Classification of real values

I have values of attribute between 0 and 1 which i want to predict. The distribution of values is shown in fig. I want to predict this attribute. The problem is there are around 15 classes in this ...
0
votes
1answer
771 views

How to represent no-detection in a confusion matrix

So let's say we have a multi-class classification problem and we want to represent the outcomes as confusion matrix. All the examples I'm finding on the web asume that all the elements are detected. I ...
0
votes
0answers
59 views

Suitable non-deep classification algorithm for binary images

I've to classify images of hand shapes like this: I've tried this methodes actually: SVM with contour vector of the hands shape as features PCA on images pixels + SVM Have you other ideas to deal ...
0
votes
1answer
70 views

Can cluster analysis of preclassified items gives idea about the classification performance?

Suppose in a classification we have a dataset with many features and their class, we want to select some features using which we can construct a classifier. We perform the cluster evaluation for the ...
0
votes
2answers
307 views

Non-linear transformation to increase separability between clusters

I want to do a classification on PC scores. I have a $400$ dimensional matrix, e.g. $2000\times 400$ ($2000$ number of samples and $400$ dimensions). I first apply PCA on it and take it to 3D, i.e. $...
0
votes
0answers
41 views

How to build a model that maps strings to lists of strings?

I have a mapping from strings to rows in a data table. Each strings maps to exactly one row in a table but the opposite is not generally true (one row can be "bound" to different strings. For example ...
0
votes
1answer
297 views

Bayes decision rule and thresholding

The best possible classification is for a set of samples drawn from any probability distribution is given by the Bayes decision rule. For any distribution, the rule is given by $$ f(x) = 1 \quad\...
0
votes
2answers
46 views

Quantify quality of multi label assignment

I am interested in quantifying how well a multi label assignment performs. E.g. given 3 coloured boxes red, green and blue, with 20 likewise coloured balls in each. A monkey is handed all the balls ...
0
votes
0answers
286 views

classification for concordance index

In this paper (and in most of the others I found), the authors want to find a continuous predictor that maximizes the concordance index. I would like to have a binary classifier instead that ...
0
votes
1answer
72 views

Doubt about feature selection

I'm working on a text classification problem using Python and NLTK. I've got two frequency distributions, one for each class (it's basically a binary classification). So, my doubt it's if there's a ...
0
votes
1answer
87 views

How to decide if a classifier is linear or not?

If the decision boundary of a binary classifier consists of multiple hyperplanes, is it still a linear classifier? If not, in multi-class classification, how do we define linear classifier? Can we ...
0
votes
0answers
121 views

classification on time series data

This is a problem related to classification on a time-ordered sequence of events. I have a data set that consists on a map between two types, A and B, of user identifiers, so that the map is one-to-...
0
votes
0answers
75 views

Does it make sense to compare on a dummy model?

Say someone were to construct a binary classifier A on some variables. They want to determine whether those variables help with prediction or not. So they build a second binary classifier which ...
0
votes
1answer
47 views

Binary random variable, big data frame: does my approach make sense?

I have a large data frame with about 1100 columns containing integers and about 30'000 rows. The last column contains a binary random variable which attains values 0 and 1. 30% of the data frame ...
0
votes
1answer
360 views

How to measure classifier performance on small and skewed sample dataset?

I have a small sample dataset (n=25) that represents the ground truth for a larger set (n=10k). I am doing a classification task and obtain, say, 3 true positives, 20 true negatives, 1 false positive, ...
0
votes
1answer
82 views

Prediction without labelled data

I am working on a churn prediction model, where I am trying to predict probability of employee churn. For each employee I have the following features 1) Role 2) Total experience 3) Current experience ...
0
votes
1answer
53 views

Classifier performance difference

I am using SVM and Random forest for classification purpose on a dataset. I am able to optimise the SVM parameters and SVM is providing very good performance in terms of accuracy, recall. But,at the ...
0
votes
1answer
42 views

Using Clustering Coefficient to Improve Naive Bayesian Classifier

I am new at statistics and ML. Due to my lack of theoretical background I was wandering if does it make sense to combine NBC and CC. I am participating to the kaggle competition https://www.kaggle.com/...
0
votes
0answers
188 views

Find the specific hypothesis in binary classification problem

Let $$A+ = \{(1,0), (-1,0), (0,1), (0,-1)\}$$ and $$A- = \{(2,2), (-2,2), (2,-2), (-2,-2)\}$$ represent the positive and negative training instances respectively. The hypothesis space used in this ...
0
votes
0answers
37 views

Is there a package in R allowing to test whether a ROC curve is dominating the other?

Among other measures, I'd like to compare ROC curves for 2 classification methods. Is there a way to test automatically if one classifier performs better independent of the threshold? Is there a ...
0
votes
0answers
25 views

evaluating an intervention against a predictive model

I have the following situation: a large sample of the population, based on which I build a predictive classification model, about a likelihood of a certain event to happen to certain persons in a ...
0
votes
0answers
1k views

Accuracy decreased after feature selection

For my machine learning study, I tested different algorithms like SVM, SMO, Naive Bayes, Trees etc. All the algorithms resulted with low accuracy levels. In fact the highest accuracy I obtained was 46%...
0
votes
1answer
27 views

Voting result when majority points to unclassified

What is common approach when dealing with classifier ensembles and majorityvoting when majority of classifier points to "unclassified". Lets say, We have ensemble consisting of 5 classifiers, 4 does ...
0
votes
0answers
2k views

Cutoff value in linear discriminant analysis with two groups

I have a simple linear discrimininant analysis with two classes. Prior probabilitiest are fixed to 0.5 and number of cases is equal between groups. In this case the cutoff value could be calculated as ...
0
votes
0answers
70 views

Predict class label at time t given time-series of vectors up to time t

Given a series of $k=4$ vectors (in $\mathbb{R}^n$), with $n=70$ at time $t=-3$, ..., $t=0$, and class labels for vectors at $t=-3$, ..., $t=-1$. Which machine learning approach would be best for ...
0
votes
1answer
394 views

How to build classification model towards some rare response classes?

I was asked to build a predictive classification model that can predict some types of response. I am interested in 6 classes, however, the total occurence of these 6 classes (out of almost half a ...
0
votes
1answer
217 views

Calculate thresholds of factor analysis output to classify data to 5 classes

Suppose that we calculate a composite indicator for some companies using Factor Analysis (FA) by combining five features to one output (calculate weights of input features). This is histogram of ...
0
votes
0answers
42 views

Why would one combine probabilties using a sqrt(-log) form?

I've been using a value, from a paper, combining two probabilities $P_A$ (for the probability that an event is A) and $P_B$ (for the probability that an event is B), where A/B are mutually exclusive ...
0
votes
0answers
122 views

Models that can learn incrementally

I want to implement a multinomial text classifier for my 3 classes 'A', 'B' and 'C'. This classifier should be open, in the sense that it should be extensible with new data. Using the hashing trick it ...
0
votes
0answers
157 views

What are the reasons why a classifier could produce bad results?

I know of four possible reasons: overfitting underfitting input data doesn't represent the problem (which I guess is underfitting) classifier isn't suitable (e.g. problem is not linear) Are there ...
0
votes
0answers
48 views

Grouping labelled object, not unsupervised clustering: Any method? Hidden markov models?

Say I have many articles, with the following labels. "classical music","pop music","physics","chemistry". Then an obvious way to group these labelled data is to group the former two under the "music" ...
0
votes
0answers
370 views

Interpretation of results regarding ROC plot on training a classifier with caret and randomForests R packages

@Dear People, i used firstly the function train() from caret package, to construct-train a classifier with random forests on a merged microarray dataset regarding selected genes, for a binary outcome(...
0
votes
0answers
140 views

What are the best practices and methods to classify a time-dependent variable?

Imagine we are collecting some feature values (like temperature, pressure etc.) every now and then and we also record the status of an equipment (which could be healthy or faulty); like below. Status ...
0
votes
0answers
48 views

How do I estimate the acceptable range?

I need to classify candidates as "good" or "bad" where a "good" candidate's age and income, say, must fall within some ranges; I'm trying to discover the boundaries for those ranges. A candidate must ...
0
votes
1answer
46 views

Finding object in the image

I'm working on building a classifier that needs to find one particular object in the photo. I'm planning on using SIFT/SURF + kmeans for feature extraction and logistic regression for classification. ...
0
votes
0answers
26 views

I have two classes and several characteristics. Is there a way to build a profile of the typical observation of one class?

I have thousands of observations and 20+ characteristics (way more if you transform them from categorical to binary characteristics). Is there some method that can be used to build a profile of the ...
0
votes
0answers
84 views

Name for a data preparation technique for classification

Say I am looking to classify an object based on certain patterns of events that occur. If I want to use classification, one easy way to do it is to divide the time up into some unit (like per day), ...
0
votes
0answers
81 views

What are the problems of these decision trees outputs?

I have a report that using two structures of decision trees for a multi-class (4-classes) classification problem with 7 inputs. First output is for ...