com.hankcs.hanlp.model.hmm

Class SecondOrderHiddenMarkovModel

java.lang.Object

com.hankcs.hanlp.model.hmm.HiddenMarkovModel

com.hankcs.hanlp.model.hmm.SecondOrderHiddenMarkovModel

public class SecondOrderHiddenMarkovModel
extends HiddenMarkovModel

Author:

hankcs

Field Summary

Fields

Modifier and Type	Field and Description
float[][][]	transition_probability2 状态转移概率矩阵

Fields inherited from class com.hankcs.hanlp.model.hmm.HiddenMarkovModel

emission_probability, start_probability, transition_probability

Constructor Summary

Constructors

Constructor and Description
SecondOrderHiddenMarkovModel()
SecondOrderHiddenMarkovModel(float[] start_probability, float[][] transition_probability, float[][] emission_probability, float[][][] transition_probability2)

Method Summary

Modifier and Type	Method and Description
protected void	estimateTransitionProbability(Collection<int[][]> samples, int max_state) 利用极大似然估计转移概率
int[][]	generate(int length) 生成样本序列
float	predict(int[] observation, int[] state) 预测（维特比算法）
boolean	similar(HiddenMarkovModel model)
protected void	toLog()
void	unLog()

Methods inherited from class com.hankcs.hanlp.model.hmm.HiddenMarkovModel

deepCopy, drawFrom, estimateEmissionProbability, estimateStartProbability, generate, logToCdf, logToCdf, normalize, predict, similar, train

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

transition_probability2

public float[][][] transition_probability2

状态转移概率矩阵

Constructor Detail

SecondOrderHiddenMarkovModel

public SecondOrderHiddenMarkovModel(float[] start_probability,
                                    float[][] transition_probability,
                                    float[][] emission_probability,
                                    float[][][] transition_probability2)

SecondOrderHiddenMarkovModel

public SecondOrderHiddenMarkovModel()

Method Detail

estimateTransitionProbability

protected void estimateTransitionProbability(Collection<int[][]> samples,
                                             int max_state)

Description copied from class: HiddenMarkovModel

利用极大似然估计转移概率

Overrides:

estimateTransitionProbability in class HiddenMarkovModel

Parameters:

samples - 训练样本集

max_state - 状态的最大下标，等于N-1

generate

public int[][] generate(int length)

Description copied from class: HiddenMarkovModel

生成样本序列

Specified by:

generate in class HiddenMarkovModel

Parameters:

length - 序列长度

Returns:

序列

toLog

protected void toLog()

Overrides:

toLog in class HiddenMarkovModel

predict

public float predict(int[] observation,
                     int[] state)

Description copied from class: HiddenMarkovModel

预测（维特比算法）

Specified by:

predict in class HiddenMarkovModel

Parameters:

observation - 观测序列

state - 预测状态序列（需预先分配内存）

Returns:

概率的对数，可利用 (float) Math.exp(maxScore) 还原

unLog

public void unLog()

Overrides:

unLog in class HiddenMarkovModel

similar

public boolean similar(HiddenMarkovModel model)

Overrides:

similar in class HiddenMarkovModel