Хммучиться МультиномиальныйHMM Размер эмиссиипроб_ ⇐ Python

[Гость]

I'm using the code below to fit an HMM model with two hidden states, a vocabulary of size 5 (so 5 possible symbols), and a list of sequences, each with 10 observations.
I don't understand why

Код: Выделить всё

model.emissionprob_

has size

Код: Выделить всё

(2, 10)

, when the number of columns should be equal to the number of symbols available in the vocabulary, 5 in this case.

Код: Выделить всё

from hmmlearn import hmm
import numpy as np
import pandas as pd

# Define the sequences
sequences = np.random.randint(1, 6, size=(100000, 10)).tolist()

# Convert sequences to numpy array
sequences_np = np.array(sequences)

# Create and fit the Multinomial HMM model with 2 hidden states
model = hmm.MultinomialHMM(n_components=2)
model.fit(sequences_np)

# Print the model parameters
print("Initial state distribution:")
print(model.startprob_)
print("\nTransition matrix:")
print(model.transmat_)
print("\nEmission probabilities:")
print(model.emissionprob_)

Источник: https://stackoverflow.com/questions/781 ... nprob-size