I remember that several years ago (might be decades...) Lisp was the most extended language in AI applications. Around 90% of all applications were programmed, at least partially, using Lisp or some kind of Lisp-like language. Even despite of the fact that Prolog was more popular in Europe (probably, in part, because of Marseilles and Edinburgh research groups activities; in the USA, for several historical reasons, it was not so popular).
What is the current state of affairs in AI applications? Is there any language to be considered the most applied in practical creations? I understand that most of the current real-world systems are integrated by several software systems, but anyway, people continue programming using Java, Python, C#, PHP, etc., as the main component of their applications.
Python is probably the most widely used language bside c++.
There are lots of AI libraries with both c++ and python bindings. It is faster and easier to prototype in python and if needed port your model to c++ if the library support both. In many cases you don't need to port the model and simply let it run in python and serve it via API using Flask for example
Python is probably the most widely used language bside c++.
There are lots of AI libraries with both c++ and python bindings. It is faster and easier to prototype in python and if needed port your model to c++ if the library support both. In many cases you don't need to port the model and simply let it run in python and serve it via API using Flask for example
I would say R and Python specifically for start ups. When it comes to big data, Scala.
I've used prolog (BNR Prolog, SWI Prolog, YAP) for constraint-based modeling at work and home for about 30 years, primarily for small projects. It has rule-based committed-choice constraint propagation (CHR) integrated nicely with prolog's backtracking -- I've developed some scheduling apps using it. Check out the new language picat (picat-lang.org) for more mainstream capabilities. For larger projects, C++ with library support (ILOG, now owned by IBM) was expensive but fast. If you are using "AI" to mean machine learning, then python seems to the logical choice.
one of the most used programming language in AI is python and if you want to use it for machine learning spatial for hot research topics like deep learning, your best choice can be cafe and tensor flow and keras library.
R is more good for AI in terms of data analysis and visualizations. For writing algorithms and models python would be the better choice.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Artificial Intelligence