We have trained a neural network to predict the next seven days of Bitcoin prices using the preceding 76 weeks of prices. On average, that model issues predictions that are about 8.4 percent distant from real Bitcoin prices.
This section describes common strategies for improving the performance of neural network models:
- Adding or removing layers and changing the number of nodes
- Increasing or decreasing the number of training epochs
- Experimenting with different activation functions
- Using different regularization strategies
We will evaluate each modification using the same active learning environment developed by the end of the Model Evaluation section, measuring how each one of these strategies may help us develop a more precise model.
