3 ms·
One point against this algorithm is the generic downside of all machine learning techniques - if you apply your algorithm and don't get a satisfactory result ..
by roenxi 5y ago
One point against this algorithm is the generic downside of all machine learning techniques - if you apply your algorithm and don't get a satisfactory result ...
- The algorithm not amenable to this problem
- You have a buggy implementation
Which is true? This is one of the issues that genetic algorithms run into as a similar powerful and chronically underutilised tool.
Being nondeterministic isn't the end of the world, but the payoffs have to be much higher than using simple deterministic methods.