ESN学习笔记——easyesn（3）回归案例

文章目录

• 回归案例
• 训练模型
• 模型评价，测试集上均方误差

回归案例

载入包

# some_file.py
import sys
sys.path.insert(0, '../src/easyesn/')
import matplotlib.pyplot as plt
import numpy as np
from easyesn import RegressionESN
from easyesn import helper as hlp

训练集和测试集

设置输入的维度，长度

inputLength = 2000
n = 500

生成数据

x = np.empty((n, inputLength))
y = np.empty((n, 1))
domain = np.linspace(0, 4*np.pi, inputLength)
for i in range(n):
    y[i, 0] = np.random.rand()*10
    x[i] = y[i, 0]*2*np.sin(domain)

划分训练集和测试集

from sklearn.model_selection import train_test_split
x_train, x_test, y_train, y_test = train_test_split(x, y, test_size=0.33, random_state=42)

训练模型

esn = RegressionESN(n_input=1, n_output=1, n_reservoir=50, leakingRate=0.2, regressionParameters=[1e-2], solver="lsqr")

拟合

esn.fit(x_train, y_train, transientTime=100, verbose=1)

预测

y_test_pred = esn.predict(x_test, transientTime=100, verbose=1)

预测结果和真实值可视化

plt.plot(y_test_pred, "--")
plt.plot(y_test, ":")
plt.show()

模型评价，测试集上均方误差

print(np.mean((y_test_pred-y_test)**2))
print(np.mean(np.abs(y_test_pred-y_test)))

本文地址：https://blog.csdn.net/todingdong/article/details/111937624