3 ms·
This is super exciting. Because of this, I started looking into Spohn's paper as well as yours. I need to finish reading your paper, and then I'd love to ask ho
by greglearns 10y ago
This is super exciting. Because of this, I started looking into Spohn's paper as well as yours. I need to finish reading your paper, and then I'd love to ask how you went from the theory to the algorithm. I'd love to understand what would be involved with porting it to a functional language (Rust most likely, which could help you in terms of making it embeddable in most any other language, for those that do not use Java).
Thank you for your great work!
- tr352 10y agoThanks! The principle could be applied to many languages, just like probabilistic programming also has been applied to many languages (look for example at Venture, Church, IBAL). This is something I'd like to do in the future. Perhaps the easiest approach is to try to integrate the ranking semantics in an existing probabilistic language such as Church. Please contact me if you have any questions, suggestions, ideas etc.