What are the best practices for selecting a random state in train test split when evaluating the performance of cryptocurrency trading strategies?

1 answers

• Donna UpchurchJun 11, 2021 · 4 years ago
  When it comes to selecting a random state in train test split for evaluating cryptocurrency trading strategies, there are a few best practices to keep in mind. First, using a fixed seed value ensures reproducibility, allowing you to compare the performance of different strategies accurately. Second, it's important to choose a random state that represents the overall dataset well. This can be achieved by using a random state that is a good representation of the dataset's characteristics, such as the mean or median value of a relevant feature. Finally, it's worth considering the distribution of the dataset when selecting the random state. If the dataset has a skewed distribution, it may be beneficial to use a random state that is closer to the mean or median value. By following these best practices, you can ensure a fair evaluation of cryptocurrency trading strategies in train test split.
  899

  450