Hands-On Machine Learning for Cybersecurity

Autocorrelation

In order to choose two variables as a candidate for time series modeling, we are required to perform a statistical correlation analysis between the said variables. Here each variable, the Gaussian curve, and Pearson's coefficient are used to identify the correlation that exists between two variables.

In time series analysis, the autocorrelation measures historic data called lags. An autocorrelation function (ACF) is used to plot such correlations with respect to lag. In Python the autocorrelation function is computed as follows:

import matplotlib.pyplot as pltimport numpy as npimport pandas as pfrom statsmodels.graphics.tsaplots import plot_acfdata = p.Series(0.7 * np.random.rand(1000) + 0.3 * np.sin(np.linspace(-9 * np.pi, ...