遗传优化算法优化LSTM-MSE

话不多说，直接上代码

数据

实验数据格式为单列的时间序列数据，不方便给出

神经网络部分

#!usr/bin/env python
# -*- coding:utf-8 _*-
"""
@author: liujie
@software: PyCharm
@file: deep_learning.py
@time: 2020/11/18 11:06
"""
import warnings
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from sklearn.preprocessing import StandardScaler,MinMaxScaler
from sklearn.metrics import mean_squared_error, mean_absolute_error

import tensorflow as tf
from tensorflow import keras
from tensorflow.keras import layers, optimizers, metrics
from keras.layers.recurrent import LSTM


# 加载数据
def read_data(filepath):
    data = pd.read_csv(filepath, index_col=0)
    data.fillna(method='pad', inplace=True)
    # data.dropna(inplace=True)
    return data


# 分割数据
def splitdata(data, windows, horizon):
    all_data = []
    for dta in range(len(data) - windows - horizon + 1):
        all_data.append(data[dta: (dta + windows)])

    all_data = np.array(all_data)
    ydata = data[horizon - 1 + windows:]
    ydata = list(ydata)

    # 构造y变量
    all_data = pd.DataFrame(all_data)
    all_data['y'] = ydata
    all_data = np.array(all_data)

    # 训练集/测试集-0.8、0.2
    row_1 = round(0.2 * len(all_data))

    x_train_initial = all_data[:-row_1, :-1]
    y_train_initial = all_data[:-row_1, -1]


    x_test_initial = all_data[-row_1:, :-1]
    y_test_initial = all_data[-row_1:, -1]

    return x_train_initial, y_train_initial, x_test_initial, y_test_initial


# 归一化
def min_max_scaler(x_train_initial, y_train_initial, x_test_initial, y_test_initial):
    mm_x = MinMaxScaler()
    x_train = mm_x.fit_transform(x_train_initial)
    x_test = mm_x.transform(x_test_initial)

    mm_y = MinMaxScaler()
    y_train = mm_y.fit_transform(y_train_initial.reshape(-1, 1))
    y_test = mm_y.transform(y_test_initial.reshape(-1, 1))

    return x_train, y_train, x_test, y_test,mm_y

def load():
    index = 'data'
    windows = 8
    horizon = 1
    filepath = '../data/Data.csv'
    data = read_data(filepath)[index]
    # 分割数据
    x_train_initial, y_train_initial, x_test_initial, y_test_initial = splitdata(data,windows,horizon)
    # 归一化
    x_train, y_train, x_test, y_test, mm_y = min_max_scaler(x_train_initial, y_train_initial, x_test_initial