测量控制领域的AI（人工智能）活用

# Install Python *Only when the Python environment is not installed
sudo apt update
sudo apt upgrade
sudo apt install build-essential libbz2-dev libdb-dev libreadline-dev libffi-dev libgdbm-dev liblzma-dev libncursesw5-dev libsqlite3-dev libssl-dev zlib1g-dev uuid-dev tk-dev
wget https://www.python.org/ftp/python/3.10.0/Python-3.10.0.tgz
tar -xf Python-3.10.0.tgz
cd Python-3.10.0/
./configure --enable-optimizations
make -j $(nproc)
sudo make altinstall

# Install library
pip install --upgrade pip
pip install tensorflow
pip install numpy pandas matplotlib scikit-learn opencv-python

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from sklearn.preprocessing import MinMaxScaler
from keras.models import Sequential
from keras.layers import Dense
from IPython.display import display

# Data window width
input_shape_num = 50

# Normalizing the dataset
scaler = MinMaxScaler()
# Calculating the transformation formula and perform data transformation
train_list = scaler.fit_transform(train_data[['Signal']], test_data[['Signal']])
test_list = scaler.transform(test_data[['Signal']])

# Flatten to one dimension array
train_list = train_list.flatten()
test_list = test_list.flatten()

# Creating partial time series
train_vec = []
test_vec = []
for i in range(len(train_list)-input_shape_num+1):
    train_vec.append(train_list[i:i+input_shape_num])
for i in range(len(test_list)-input_shape_num+1):
    test_vec.append(test_list[i:i+input_shape_num])

# Setting autoencoder
def init_autoencorder(input_shape_num):
    model = Sequential()
    # Encoding (number of output units = units, activation function = relu, number of inputs = input_shape_num)
    model.add(Dense(units=200, activation='relu',
              input_shape=(input_shape_num,)))
    # Decoding (Number of units to output = units, activation function = relu)
    model.add(Dense(units=100, activation='relu'))
    # Output layer (number of outputs = input_shape_num, activation function = sigmoid)
    model.add(Dense(input_shape_num, activation='sigmoid'))

    # Checking the model created
    model.summary()

    return model

# Running autoencoder
def run_auto_encoder(model, train_vec, batch_size, epochs):
    # Set learning conditions (optimization = adam method, loss function = mean squared error, evaluation function = accuracy rate of multi-class classification)
    model.compile(optimizer='adam', loss='mse', metrics=['acc'])

# Executing learning (training data = train_vec, teacher data = train_vec, number of gradient update samples = batch_size,
    hist = model.fit(x=train_vec, y=train_vec, batch_size=batch_size, number of training data iterations = epochs, progress display = progress bar, training data ratio)
                     epochs=epochs, verbose=1, validation_split=0.2)

    return hist

# Evaluating the model on test data
pred_vec = model.predict(test_vec)
pred_vec = scaler.inverse_transform(pred_vec)
test_vec = scaler.inverse_transform(test_vec)

# Outputting test data to graph
plt.figure(figsize=(16, 2))
plt.plot(test_vec[:, 0], clor='blue', label='Test')
plt.plot(pred_vec[:, 0], color='orange', label='Filter')
plt.legend()
plt.show()