Implementing reinforcement learning in Python

Let's now move on to an example in which streaming data is used for Q-Learning. The data that we will be using is simulated data of stock prices:

1. The data is generated in the following block of code.

The list of values that is first generated is a list of 30,000 consecutive values that represent stock prices. The data generating process is the starting point of 0 and at every time step, there is a random value added to this. The random normal values are centered around 0, which indicates that prices would go up or down by a step size based on a standard deviation of 1.

This process is often referred to as a random walk and it can go quite far up or down. After that, the values are standardized to be within a normal distribution again.

Code Block 8-1

import numpy as np
import matplotlib.pyplot as plt
import random
starting = 0
values = [starting]
for i in range(30000):
    values.append(values...