What is the fundamental reason that statistics-based AI (e.g., ML and Neural Net) has become more popular than other forms of AI, e.g., Fuzzy Logic and rules-based AI (e.g., Prolog)?
$\begingroup$
$\endgroup$
1
-
$\begingroup$ Isn't statistics itself rule based from the point of view of human understanding? For example, does cross-validation count as some kind of rule? $\endgroup$– cinchCommented Nov 27, 2022 at 4:21
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
2
The availability of large data sets.
In symbolic/rule-based AI, the 'knowledge' has to be hand-coded, usually by experts. This is expensive and limited to small-scale problems only.
In statistical AI/Machine Learning approaches, this is replaced by training a system on data. With very large amounts of data it is possible to achieve results that are not possible with rule-based approaches. In the past, such data sets were usually not available.
However, this also means that the AI system has become somewhat of a black box, as it is not possible (at least not easily) why the system has behaved in a particular way.
So in areas where this is important (eg banking and finance), rule-based approaches are still valuable.
-
-
1$\begingroup$ Fuzzy logic is basically a different way to express rule-based information. $\endgroup$ Commented Sep 13, 2022 at 16:23