|
|
|
|
|
|
|
|
|
Irreflexive
| Глоссарий по искусственному интеллекту |
An irreflexive binary relationship r is one in which the relationship a ra is not true. an example would be the relationship "is a parent of". see also: partially ordered set, reflexive.
|
|
Relationship, английский
- Отношение: соотношение; связь; взаимоотношение; взаимосвязь
- Отношение, степень родства
- A way in which someone or something is connected to another the incidence of the disease has a close relationship to the environment. he became withdrawn and broke off all relationships with his family.
- Зависимость, соотношение
- A connection between objects.
- A logical connection between entities.
- A smartart graphic layout type that includes layouts designed to illustrate connections.
- An association established between common fields (columns) in two tables. a relationship can be one-to-one, many-to-many, or one-to-many.
|
Interval attribute, английский
An interval valued attribute uses numeric values for which the relative differences are meaningful whose meaning would not be changed if the values were translated or multiplied by a positive constant. the zero in these scales is arbitrary. a familiar example would be temperature in degrees celsius. the difference between 20 and 30 is twice as large as the difference between 20 and 25, but the 30-degree temperature is not 50 percent larger than the 20-degree temperature. the zero value in this scale is arbitrary. this level of attribute supports the same operations as do ordinal and nominal value attributes as well as sums and differences. scale and origin dependent operations, such as multiplication and exponentiation, do not have an intrinsic meaning for these type of attributes. see also: machine learning, nominal (attribute type), ordinal attribute, ratio attribute.
Occam algorithm, английский
An occam algorithm is a general structure for applying probably approximately correct (pac) models. it has two basic steps. first, draw a random sample from the target distribution, and then return all the rules that are consistent with the concepts. the size of the sample should be large with respect to the number of attribute and the set of possible rules.
|
|
|
|
|
|
|
