基于多头注意力机制LSTM股价预测模型

1、多头注意力机制层的构建

class MultiHeadAttention(tf.keras.layers.Layer):
  def __init__(self, num_heads, d_model):
    super(MultiHeadAttention, self).__init__()
    self.num_heads = num_heads
    self.d_model = d_model

    assert d_model % self.num_heads == 0

    self.depth = d_model // self.num_heads

    self.wq = tf.keras.layers.Dense(d_model)
    self.wk = tf.keras.layers.Dense(d_model)
    self.wv = tf.keras.layers.Dense(d_model)

    self.dense = tf.keras.layers.Dense(d_model)

  def split_heads(self, x, batch_size):
    """Split the last dimension into (num_heads, depth).
    Transpose the result such that the shape is (batch_size, num_heads, seq_len, depth)
    """
    x = tf.reshape(x, (batch_size, -1, self.num_heads, self.depth))
    return tf.transpose(x, perm=[0, 2, 1, 3])

  def scaled_dot_product_attention(self, q, k, v, mask):
    """Calculate the attention weights.
    q, k, v must have matching leading dimensions.
    k, v must have matching penultimate dimension, i.e.: seq_len_k = seq_len_v.
    The mask has different shapes depending on its type(padding or look ahead) 
    but it must be broadcastable for addition.

    Args:
      q: query shape == (..., seq_len_q, depth)
      k: key shape == (..., seq_len_k, depth)
      v: value shape == (..., seq_len_v, depth_v)
      mask: Float tensor with shape broadcastable 
            to (..., seq_len_q, seq_len_k). Defaults to None.

    Returns:
      output, attention_weights
    """

    matmul_qk = tf.matmul(q, k, transpose_b=True)  # (..., seq_len_q, seq_len_k)

    # scale matmul_q
    # scale matmul_qk
    dk = tf.cast(tf.shape(k)[-1], tf.float32)
    scaled_attention_logits = matmul_qk / tf.math.sqrt(dk)

    # add the mask to the scaled tensor.
    if mask is not None:
      scaled_attention_logits += (mask * -1e9)  

    # softmax is normalized on the last axis (seq_len_k) so that the scores
    # add up to 1.
    attention_weights = tf.nn.softmax(scaled_attention_logits, axis=-1)  # (..., seq_len_q, seq_len_k)

    output = tf.matmul(attention_weights, v)  # (..., seq_len_q, depth_v)

    return output, attention_weights

  def call(self, v, k, q, mask):
    batch_size = tf.shape(q)[0]

    q = self.wq(q)  # (batch_size, seq_len, d_model)
    k = self.wk(k)  # (batch_size, seq_len, d_model)
    v = self.wv(v)  # (batch_size, seq_len, d_
    # split heads
    q = self.split_heads(q, batch_size)  # (batch_size, num_heads, seq_len_q, depth)
    k = self.split_heads(k, batch_size)  # (batch_size, num_heads, seq_len_k, depth)
    v = self.split_heads(v, batch_size)  # (batch_size, num_heads, seq_len_v, depth)

    # scaled dot product attention
    scaled_attention, attention_weights = self.scaled_dot_product_attention(q, k, v, mask)

    # concatenation of heads
    scaled_attention = tf.transpose(scaled_attention, perm=[0, 2, 1, 3])  # (batch_size, seq_len_q, num_heads, depth)
    concat_attention = tf.reshape(scaled_attention, 
                                  (batch_size, -1, self.d_model))  # (batch_size, seq_len_q, d_model)

    # final linear layer
    output = self.dense(concat_attention)  # (batch_size, seq_len_q, d_model)

    return output

2、构建股价预测模型

# Stock price prediction model
class StockPricePredictionModel(tf.keras.Model):
  def __init__(self, num_heads, d_model, num_lstm_units):
    super(StockPricePredictionModel, self).__init__()
    self.num_heads = num_heads
    self.f_model = d_model
    self.num_lstm_units = num_lstm_units

    self.multi_head_attention = MultiHeadAttention(self.num_heads, self.d_model)
    self.lstm = tf.keras.layers.LSTM(self.num_lstm_units, return_sequences=True)
    self.dense = tf.keras.layers.Dense(1)

  def call(self, inputs, mask):
    attention_output = self.multi_head_attention(inputs, input, input, mask)
    lstm_output = self.lstm(attention_output)
    prediction = self.dense(lstm_output)
    return prediction

model = StockPricePredictionModel(num_heads=9, d_model=256, num_lstm_units=128)

3、模型训练与结果