Predicts the most suitable category label for given textual data. More...
Inheritance diagram for Classifier:
Public Member Functions | |
| train (array &$dataTrain, array &$dataLabel) | |
| Trains the classifier with the given example data. | |
| setDatabase ($name) | |
| Sets the name of the database to operate on. | |
| save () | |
| Uploads the learnt parameters to a new database. | |
| classify ($dataTest) | |
| Predicts the most suitable category label for the given test data. | |
Data Fields | |
| const | OOV = "__OOV__" |
| Out-of-vocabulary common symbol. | |
Detailed Description
Predicts the most suitable category label for given textual data.
Definition at line 44 of file Classifier.php.
Member Function Documentation
| Classifier::classify | ( | $ | dataTest | ) |
Predicts the most suitable category label for the given test data.
- Parameters:
-
dataTest The test data, e.g., in textual form.
- Returns:
- The predicted category label.
- Exceptions:
-
Exception if the connection to the working database cannot be established.
- Precondition:
- The test data should be a valid string.
- A trained model should be available on the working database, i.e., an existing and non-empty database.
- Postcondition:
- The most appropriate category label is delivered.
Implemented in MultinomialNaiveBayes.
| Classifier::save | ( | ) |
Uploads the learnt parameters to a new database.
- Exceptions:
-
Exception if the connection to the database management system fails.
- Precondition:
- The classifier should be trained in computer memory.
- A valid and original database name should be set.
- Postcondition:
- The learnt parameters are saved on the new database.
Implemented in MultinomialNaiveBayes.
| Classifier::setDatabase | ( | $ | name | ) |
Sets the name of the database to operate on.
- Parameters:
-
name Name of the working database.
- Precondition:
- The given name should be a valid string.
- Postcondition:
- The name of the working database is set.
Implemented in MultinomialNaiveBayes.
| Classifier::train | ( | array &$ | dataTrain, | |
| array &$ | dataLabel | |||
| ) |
Trains the classifier with the given example data.
- Parameters:
-
dataTrain Data instances, e.g., in textual form. dataLabel Instance labels.
- Exceptions:
-
Exception if data/label sizes don't match, i.e., different numbers of labelled instances are given.
- Precondition:
- Parameters must be arrays of the same length.
- Postcondition:
- Classification model is learnt in computer memory.
Implemented in MultinomialNaiveBayes.
Field Documentation
| const Classifier::OOV = "__OOV__" |
Out-of-vocabulary common symbol.
Definition at line 49 of file Classifier.php.
Referenced by MultinomialNaiveBayes::classify(), and MultinomialNaiveBayes::train().
The documentation for this interface was generated from the following file:
