User:Francis Tyers/Perceptron

A perceptron is a classifier that

The classifier consists of:

• Binary features
• Weights

Example

Here is a worked example of a perceptron applied to the task of lexical selection. Lexical selection is the task of choosing a target translation $t*$ for a given source word $s$ in a context $C$ out of a set of possible translations $T$. A perceptron makes a classification decision for a single class, so we need to train a separate perceptron for each possible target word selection.

In the example,

• $s$ = estació
• $T$ = {season, station}
• $t*$ = season

Features

The features we will be working with are ngram contexts around the "problem word". These can be extracted from the word alignments calculated from a parallel corpus.

Catalan	English
Durant l' estació seca les pluges són escasses.	During the dry season it rains infrequently.

Training data

_ sec	1
_ de el any	1
_ de tren	0
_ de el línia	0
_ humit	1
_ plujós	1
un _ a	0