Disliked{quote} ok...you post make sense and I have little knowledge in the field... but you also have to consider that no machine can compare with my creativity... my rules are based on my creativity and my creativity has more adaptability then a machine, machine needs to be coded (correct me if I am wrong) where I also can bend my rules to adapt to the actual imminent changes of the conditions, I can also rely on my gut feeling which I use often always combined by my rules.Ignored
If you constantly add new, untested and unproven rules into your system then you essentially create randomness. How do you know that the new rules will lead to positive outcome? This is not how profitable trading works!
The only way to find real edge in trading is through statistical analysis! Backtesting and Forward testing.
The standard process goes like this:
1. You come up with some idea! Maybe you spotted some repeating pattern or whatever.....
2. You define the pattern with maximum precision.
3. You build different rules based on the pattern properties in attempt to extract positive expectancy.
4. You test the rules with historical data to see if they work.
5. You optimize the parameters of the rules to find what works best under different conditions.
6. After all testing is complete, and if you found positive expectancy then you can try the model in real time to see if it works the same way with live data.
It's a long process!
But all of these steps from 1 to 6 can be automated! It's possible to create a robot to write it's own code and to search for patterns on it's own. It's just very difficult. There are some paid applications which can do all of this, but they are still quite primitive. StrategyQuant is one such example.
1
7