Classifier

(redirected from classifiers)
Also found in: Dictionary, Thesaurus.

classifier

[′klas·ə‚fī·ər]
(mechanical engineering)
Any apparatus for separating mixtures of materials into their constituents according to size and density.

Classifier

 

(in mineral concentration), an apparatus for separating mixtures of mineral particles into fractions according to size, shape, or density. Depending upon the medium in which separation of the materials occurs, classifiers are divided into hydraulic and pneumatic. Depending upon the force used, there are gravity classifiers, centrifugal classifiers, and high-tension separators.

Wet mechanical classifiers used for preparing ores for flotation are the most common. Drag tanks and spiral classifiers (in which the settled material is removed by rakes or spirals similar to Archimedes’ screw) are usually used for this purpose. The drag tank is an inclined trough with the suspension (pulp) being delivered into the lower third of the trough. The settled “sands” are moved up along the bottom of the trough by the reciprocating motion of the rakes along an ellipsoid path; in addition, the “sands” are freed of water. The fine particles which do not have time to settle are removed with the “overflow.” The fineness of the overflow depends on the rocking rate of the rakes and the incline of the trough. The pulp density (the mass content of the solid phase in the pulp) is the most important method for the operational control of the classifier. With an increase in the density, the size of the particles passing into the overflow increases (since the speed at which they fall decreases in a viscous, dense medium). But with a very great dilution of the pulp the size of the particles in the overflow also increases, since there is an increase in the rate of flow of the upward water stream which removes all the particles, including the large ones. The principle of operation of spiral classifiers is the same; they merely have a simpler and more dependable design.

A mechanical classifier is most often used together with ball and rod mills in crushing ore for the continuous extraction of rather fine particles from the material being crushed. An important advantage in the design of such classifiers is the raising of the sands above the feed entry point, as this makes it possible to carry out closed-circuit crushing easily. The productivity of the mechanical classifiers depends upon the width of the overflow. For this reason simplex, duplex, and quadruplex classifiers are used (with an overflow width of from 200 to 3,000 mm), as well as a single-, double-, and four-spiral classifiers (with a width from 300 to 2,400 mm). Conical or pyramidal hydraulic classifiers without moving parts were also widely used until the beginning of the 1950’s. They are suitable primarily for the precision classification of small quantities of the finest products. Hydrocyclones and centrifuges are types of centrifugal classifiers.

Pneumatic classifiers are subdivided into chamber and centrifugal classifiers, with the latter being used more often. They consist of a blower, a distributing plate, and a centrifugal bladed wheel which spins the large particles toward the walls of the inner cone. These particles are carried by the airstream and settle in the inner cone. The fine fraction settles in the outer cone where the velocity of the airstream is low. There are many modifications of pneumatic classifiers.

REFERENCES

See references under CLASSIFICATION.

V. I. KLASSEN


Classifier

 

(also called numerative), an auxiliary lexeme or noun that has lost its basic meaning to a greater or lesser degree and is used to designate countable objects. Classifiers are used in an attributive word group that contains a numeral and a noun; an example in Russian is piat’ shtuk karandashei (“five pencils”; literally, “five pieces of pencils”). Classifiers exist in many languages of Eastern and Southeastern Asia, such as Chinese, Vietnamese, and Indonesian, and in the Indo-Iranian, Turkic, and Dravidian languages.

Classifiers generally indicate the semantic class to which a noun belongs: compare Tadzhik du nafar korgar (“two workers”), du sar gusfand (“two sheep”), and du dona seb (“two apples”). Suffixal elements may fulfill the same function: compare Tadzhik duta odam (“two people”) and duta kitob (“two books”).

In some languages, classifiers also function as particles indicating singleness and indefiniteness: compare Oriya mote khaṇḍe pustaka die (“give me some [any] book”) and cāri khaṇḍe āmba (“four mangoes”; literally, “four pieces of mangoes”). The words khaṇḍe and khaṇḍa are particles and classifiers for a class of nouns designating round, flat, and oblong objects.

D. I. EDEL’MAN

References in periodicals archive ?
As in the case of adversarial data generation, research in transferring adversarial examples has focused mainly on what characteristics of learner models of classifiers favor transferring adversarial examples across the models.
Human baseline is another important factor in estimating the performance of artificial classifiers. We hire 20 human subjects whose ages are in twenties and thirties and arrange them in two groups: a training group and a non-training group.
This classifier assesses specific metabolites, types of bacteria, functional pathways, and other factors.
Some these classifiers are lazy which store training data and wait until it is given a test tuple, usually less time in training but more time in predicting such as instance-based and kNN.
Afterwards, the inputs of the classifiers are obtained by using a feature extraction stage.
After identifying the classifiers, their performance was evaluated using sensitivity and specificity tests (Portney & Watkins, 2009) to assess the effectiveness of the prediction of IDH occurrence.
In this section, details of the dataset used in the experiments, the description of the metrics used for performance evaluation, and the report of the tested classifiers are given, as well as the experimental design conducted.
Different classifiers such as SVM, NB, and alternating decision tree used for review classification and more than 75% accuracy was achieved.
In the case of underwater acoustic target classifiers, when the signal spectrum is superimposed by a noise spectrum additively, it can be safely assumed that the spectrum of the collected signal is a linear sum of the signal spectrum and the noise spectrum [32].
To validate ELRDD, SVM, GMM, and LR were compared as classifiers for detecting depression.
This learning framework consists of two main stages: selecting base classifiers and evaluating rebalancing strategies.
(NASDAQ: VCYT) has presented data demonstrating the accuracy of identification of BRAF V600E and parathyroid tissue through new classifiers developed in conjunction with the next-generation Afirma Genomic Sequencing Classifier, the company said.