The traditional optimizer has been obsolete.
# The traditional optimizer has been obsolete.
# Fintechee has some useful features to help advanced traders get a good model with appropriate parameters.
One is an algorithm based on the traditional optimizer and the second one is an algorithm based on Artificial Intelligence, so-called AI, such as Neural Network.
We all know that a trading system needs to verify the parameters before using it on the real accounts. Because there are a lot of types of combinations of parameters, the way to find a better choice is really difficult.
# Consider this scenario that you have 10 parameters to set up for your trading system.
Every parameter has a range between 10 and 100. If you use 10 as the step to generate parameters, how many combinations of the parameters do you think you will generate?
The answer is 10 x 10 x 10 ... 10, 10 billion combinations!! When you run an optimizer transversally, it will take hours, even days to find a good choice. If the choice is useful, the backtesting to find the choice is worth doing. But, in case the samples are sensitive to the backtesting period and the historical data that you pick up. How can you guarantee the choice is really helpful? Picking up a choice from so many samples is just wasting time.
So the traditional optimizer or the way to optimize by transversally polling your data on your algorithms is obsolete.
# We reserve the feature
The traditional optimizer on Fintechee, but we recommend all advanced traders get started by using the second solution that I mentioned above, Artificial Intelligence to train the model of the trading system instead of the traditional optimizer.
I haven't updated my blog for several days. I was too busy recently. I'm training my neural network and the model is based on a new algorithm. I will write blog posts to discuss how to train the neural network further in the future.
We started trading since Feb 17th, 2020. Please feel free to track our trading records.
- Fintechee WEB Trader
- Account ID: 875730
- Investor Password: 1