Intersection types and (positive) almost-sure termination
نویسندگان
چکیده
Randomized higher-order computation can be seen as being captured by a lambda calculus endowed with single algebraic operation, namely construct for binary probabilistic choice. What matters about such computations is the probability of obtaining any given result, rather than possibility or necessity it, like in (non)deterministic computation. Termination, arguably simplest kind reachability problem, spelled out at least two ways, depending on whether it talks convergence expected evaluation time, second one providing stronger guarantee. In this paper, we show that intersection types are capable precisely characterizing both notions termination inside system types: lambda-term underapproximated its type, while underlying derivation's weight gives lower bound to term's number steps normal form. Noticeably, approximations tight -- not only soundness but also completeness holds. The crucial ingredient non-idempotency, without which would impossible reason reduction necessary completely evaluate term. Besides, approximation obtain proved optimal recursion theoretically: no recursively enumerable formal do better that.
منابع مشابه
Proving Positive Almost-Sure Termination
In order to extend the modeling capabilities of rewriting systems, it is rather natural to consider that the ring of rules can be subject to some probabilistic laws. Considering rewrite rules subject to probabilities leads to numerous questions about the underlying notions and results. We focus here on the problem of termination of a set of probabilistic rewrite rules. A probabilistic rewrite s...
متن کاملProving Positive Almost Sure Termination Under Strategies
In last RTA, we introduced the notion of probabilistic rewrite systems and we gave some conditions entailing termination of those systems within a finite mean number of reduction steps. Termination was considered under arbitrary unrestricted policies. Policies correspond to strategies for non-probabilistic rewrite systems. This is often natural or more useful to restrict policies to a subclass....
متن کاملBeyond Almost-Sure Termination
The aim of this paper is to argue that models in cognitive science based on probabilistic computation should not be restricted to those procedures that almost surely (with probability 1) terminate. There are several reasons to consider nonterminating procedures as candidate components of cognitive models. One theoretical reason is that there is a perfect correspondence between the enumerable se...
متن کاملOn the Hardness of Almost-Sure Termination
This paper considers the computational hardness of computing expected outcomes and deciding (universal) (positive) almost–sure termination of probabilistic programs. It is shown that computing lower and upper bounds of expected outcomes is Σ 1– and Σ 0 2–complete, respectively. Deciding (universal) almost–sure termination as well as deciding whether the expected outcome of a program equals a gi...
متن کاملAnalyzing Expected Outcomes and (Positive) Almost–Sure Termination of Probabilistic Programs is Hard
Probabilistic programs [1] are imperative sequential programs with the ability to toss a biased coin and proceed their computations depending on the outcome of the coin toss. They are used in security to describe cryptographic constructions (such as randomized encryption) and security experiments [2], in machine learning to describe distribution functions that are analyzed using Bayesian infere...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Proceedings of the ACM on programming languages
سال: 2021
ISSN: ['2475-1421']
DOI: https://doi.org/10.1145/3434313