|
|
|
|
|
|
|
|
|
Counterpropagation
| Глоссарий по искусственному интеллекту |
A kohonen learning vector quantization (lvq) network performs classification by a nearest-neighbor classification on the codebook vectors. similarly, counterpropagation performs a nearest-neighbor regression on the codebook vectors. see also: learning vector quantization, nearest-neighbor.
|
|
Quantization, английский
- Квантование; дискретизация
- The correction of rhythms in a midi sequence.
Classification, английский
- Assigning data (i e , cases or observations) cases to one of a fixed number of possible classes (represented by a nominal output variable)
- Классификация; систематизация; засекречивание
- Классификация (упорядоченная группировка явлений и объектов; может утверждаться в качестве стандарта, в том числе национального или международного), см. nomenclature
- The work of putting references or components into order so as to be able to refer to them again and identify them easily the abo classification of blood
- Классификация
- N классификация language typological ~ типологическая классификация classificatory a классификационный matrix classifying a классифицирующий article
- The placement of an item under the correct number in the customs tariff for duty purposes. at times this procedure becomes highly complicated; it is not uncommon for importers to resort to litigation over the correct duty to be assessed by the customs on a given item
- Классификация; категоризация; сортировка
- The process of assigning a set of records from a database (observations in a dataset) into (usually) one of ``small" number of pre-specified disjoint categories. related techniques include regression, which predicts a range of values and clustering, which (typically) allows the categories to form themselves. the classification can be "fuzzy" in several senses of the word. in usual sense, the classification technique can allow a single record to belong to multiple (disjoint) categories with a probability (estimated) of being in each class. the categories can also overlap when they are developed either through a hierarchical model or through an agglomerative technique. finally, the classification can be fuzzy in the sense of using "fuzzy logic" techniques. see also: clustering, fuzzy logic, regression. classification and regression trees (cart) classification and regression trees (cart) is a particular form of decision tree used in data mining and statistics.
- Классификация. для порошков (пыли)—разделение образца на фракции по размеру, форме и плотности частиц с помощью жидкости; основано, например, на различной скорости оседания фракций в жидкости или на разном поведении частиц в потоке жидкости.
- A systematic organization of classes.
- The placement of an incident into a hierarchy of descriptors that indicate what the incident is generally about. for example, an incident could be classified as being related to software, and then to microsoft, and then to word 2003.
- The type of updates that sce downloads from microsoft update during synchronization.
- An offender is classified for a particular security level and housed in an appropriate facility based on assessment of their crime, security risk and prison behavior.
- A code which provides a method for categorizing the invention.
- The code for a specific type of complementary work or variation.
- Классификация, номенклатура
Regression, английский
- Statistical technique used to evaluate relationships among variables (22).
- Регрессия
- 1. a stage where symptoms of a disease are disappearing and the person is getting better 2. (in psychiatry) the process of returning to a mental state which existed when the person was younger
- Regression commonly refers to the process of developing an empirical (data-driven) model to predict and/or explain one or more attributes in a database or set of data. it is most frequently associated with the simple linear model (y=mx+b) taught in most introductory statistics courses; the same ideas have been extended in many directions, including classification problems. when the emphasis is on hypothesis testing and simple models, the regression output is typically a few parameters that provide a direct linkage from the input variables to the predicted variables (or classification). in other situations the emphasis is on explaining as much of the variability in the output variables as is "reasonable" from the input variables. in this case, there are a number of "advanced" techniques, such as smoothing splines, decision trees, neural nets, and so forth, for which there are many "free" parameters. the meaning of any one of these parameters can be obscure. many data mining techniques are, at their core, variations on well-known regression techniques. see also: classification, clustering, decision trees, neural nets.
- The reappearance of a previously fixed problem.
- The statistical process of predicting one or more continuous variables, such as profit or loss, based on other attributes in the dataset.
- A mathematical technique used to explain and/or predict. the general form is y = a + bx + u, where y is the variable that we are trying to predict; x is the variable that we are using to predict y, a is the intercept; b is the slope, and u is the regression residual. the a and b are chosen in a way to minimize the squared sum of the residuals. the ability to fit or explain is measured by the r-square.
- A seaward retreat of a shoreline, generally expressed as a seaward
|
Procedural representation, английский
A knowledge base that uses a procedural representation (as opposed to a declarative representation) represents knowledge as a series of procedures and associated data structures. an example would be a knowledge base that contains a "person" procedure. this procedure would contain a list of persons, such as socrates, aristotle, archie bunker, and plato. when executed with an argument of "descartes," the person procedure would search its list of persons and return false. a further procedure called "mortal" could build on this with a list of mortal things, including the persons() function. then, the call mortal("plato``) would eventually invoke the person("plato") and return true. this form of representation has the advantage that it explicitly directs the method and order of query evaluation, but it can be very difficult to change when a new fact is added. hybrid methods, such as patterndirected invocation have been developed to overcome the limits of this form of representation. see also: declarative representation, knowledge base, knowledge representation.
Kohonen network, английский
A kohonen network is a form of unsupervised learning in neural networks and is similar to a k-means cluster analysis. each of the hidden units acts as a cluster center and the error measure is the sum of square differences between the input data and the nearest cluster center. the clusters compete with each other for the right to respond to an input pattern. there are several types of kohonen networks, including vector quantization, som (self-organizing maps), and learning vector quantization. see also: kolmogorov`s axioms the russian mathematician kolmogorov provided an axiomatic basis for probability. given a space of events s, and two subsets a and b of s, a function p() is a probability if p(a) > 0 p(s) = 1 if a b = 0, then p(a b) = p(a) + p(b). the three conditions are also referred to as positivity, normalization, and finite additivity. a slightly different form of the axioms substitutes countable additivity for the latter, allowing for countable collections of sets. see also: probability.
|
|
|
|
|
|
|
