اشترك بالحزمة الذهبية واحصل على وصول غير محدود شمرا أكاديميا
تسجيل مستخدم جديدAutomatic functions, linear time and learning
90
0
0.0
(
0
)
تأليف
John Case
اسأل ChatGPT حول البحث
ﻻ يوجد ملخص باللغة العربية
The present work determines the exact nature of {em linear time computable} notions which characterise automatic functions (those whose graphs are recognised by a finite automaton). The paper also determines which type of linear time notions permit full learnability for learning in the limit of automatic classes (families of languages which are uniformly recognised by a finite automaton). In particular it is shown that a function is automatic iff there is a one-tape Turing machine with a left end which computes the function in linear time where the input before the computation and the output after the computation both start at the left end. It is known that learners realised as automatic update functions are restrictive for learning. In the present work it is shown that one can overcome the problem by providing work tapes additional to a resource-bounded base tape while keeping the update-time to be linear in the length of the largest datum seen so far. In this model, one additional such work tape provides additional learning power over the automatic learner model and two additional work tapes give full learning power. Furthermore, one can also consider additional queues or additional stacks in place of additional work tapes and for these devices, one queue or two stacks are sufficient for full learning power while one stack is insufficient.
قيم البحث
اقرأ أيضاً
Many reinforcement learning (RL) environments in practice feature enormous state spaces that may be described compactly by a factored structure, that may be modeled by Factored Markov Decision Processes (FMDPs). We present the first polynomial-time a
lgorithm for RL with FMDPs that does not rely on an oracle planner, and instead of requiring a linear transition model, only requires a linear value function with a suitable local basis with respect to the factorization. With this assumption, we can solve FMDPs in polynomial time by constructing an efficient separation oracle for convex optimization. Importantly, and in contrast to prior work, we do not assume that the transitions on various factors are independent.
A distributed protocol is typically modeled as a set of communicating processes, where each process is described as an extended state machine along with fairness assumptions, and its correctness is specified using safety and liveness requirements. De
signing correct distributed protocols is a challenging task. Aimed at simplifying this task, we allow the designer to leave some of the guards and updates to state variables in the description of extended state machines as unknown functions. The protocol completion problem then is to find interpretations for these unknown functions while guaranteeing correctness. In many distributed protocols, process behaviors are naturally symmetric, and thus, synthesized expressions are further required to obey symmetry constraints. Our counterexample-guided synthesis algorithm consists of repeatedly invoking two phases. In the first phase, candidates for unknown expressions are generated using the SMT solver Z3. This phase requires carefully orchestrating constraints to enforce the desired symmetry in read/write accesses. In the second phase, the resulting completed protocol is checked for correctness using a custom-built model checker that handles fairness assumptions, safety and liveness requirements, and exploits symmetry. When model checking fails, our tool examines a set of counterexamples to safety/liveness properties to generate constraints on unknown functions that must be satisfied by subsequent completions. For evaluation, we show that our prototype is able to automatically discover interesting missing details in distributed protocols for mutual exclusion, self stabilization, and cache coherence.
We introduce MORA, an automated tool for generating invariants of probabilistic programs. Inputs to MORA are so-called Prob-solvable loops, that is probabilistic programs with polynomial assignments over random variables and parametrized distribution
s. Combining methods from symbolic computation and statistics, MORA computes invariant properties over higher-order moments of loop variables, expressing, for example, statistical properties, such as expected values and variances, over the value distribution of loop variables.
This paper studies tree-automatic ordinals (or equivalently, well-founded linearly ordered sets) together with the ordinal addition operation +. Informally, these are ordinals such that their elements are coded by finite trees for which the linear or
der relation of the ordinal and the ordinal addition operation can be determined by tree automata. We describe an algorithm that, given two tree-automatic ordinals with the ordinal addition operation, decides if the ordinals are isomorphic.
An algebraic linear ordering is a component of the initial solution of a first-order recursion scheme over the continuous categorical algebra of countable linear orderings equipped with the sum operation and the constant 1. Due to a general Mezei-Wri
ght type result, algebraic linear orderings are exactly those isomorphic to the linear ordering of the leaves of an algebraic tree. Using Courcelles characterization of algebraic trees, we obtain the fact that a linear ordering is algebraic if and only if it can be represented as the lexicographic ordering of a deterministic context-free language. When the algebraic linear ordering is a well-ordering, its order type is an algebraic ordinal. We prove that the Hausdorff rank of any scattered algebraic linear ordering is less than $omega^omega$. It follows that the algebraic ordinals are exactly those less than $omega^{omega^omega}$.
سجل دخول لتتمكن من نشر تعليقات
التعليقات
جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
