Problem definition

As we already know, portfolio management is the continuous reallocation of funds across different multiple financial products (assets). In this work, the time is divided into equal length periods, where each period T = 30 minutes. At the beginning of each period, the trading agent reallocates the fund across different assets. The price of an asset fluctuates within a period, but four important price metrics are taken into consideration, which are good enough to characterize the price movement of an asset in the period. These price metrics are as follows:

  • Opening price
  • Highest price
  • Lowest price
  • Closing price

For a continuous market (such as our test case), the opening price of an asset in a period t is its closing price ...

Get Reinforcement Learning with TensorFlow now with the O’Reilly learning platform.

O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.