|
|
|
|
|
|
|
|
|
Pandemonium
| Глоссарий по искусственному интеллекту |
One of the earliest attempts (1959) at machine learning, using an approach similar to a neural network. lowerorder demons learned to simplify features in the input data which were passed up to higher-order demons, etc. see also: machine learning, artificial neural networks.
|
|
Artificial, английский
- Искусственный
- Made to resemble a natural material or object, for example, faux marbre.
- A искусственный intelligence, language
- Bait made of some type of metal, plastic or rubber. comes in many colors, may be scented or unscented.
|
Classifier ensembles, английский
One method of improving the performance of machine learning algorithms is to apply ensembles (e.g., groups) of classifiers to the same data. the resulting classifications from the individual classifiers are then combined using a probability or voting method. if the individual classifiers can disagree with each other, the resulting classifications can actually be more accurate than the individual classifiers. each of the individual classifiers needs to have better than a 50 percent chance of correct classifications. figure c.2 —
Word sense disambiguation, английский
One of the important sub-tasks in semantic analysis of natural language is word sense disambiguation, in which a program needs to determine which of several possible meanings of a phrase is the correct one. as an example, in the sentence "the pigs are in the pen," the word pen could mean an enclosure or a writing device. a program performing a semantic analysis would need to determine that the sentence referred to an enclosure for animals.
|
|
|
|
|
|
|
