Leslie G. Valiant

A theory of the learnable

Humans appear to be able to learn new concepts without needing to be programmed explicitly in any conventional sense. In this paper we regard learning as the phenomenon of knowledge acquisition in the absence of explicit programming. We give a precise methodology for studying this phenomenon from a computational viewpoint. It consists of choosing an appropriate information gathering mechanism, the learning protocol, and exploring the class of concepts that can be learnt using it in a reasonable (polynomial) number of steps. We find that inherent algorithmic complexity appears to set serious limits to the range of concepts that can be so learnt. The methodology and results suggest concrete principles for designing realistic learning systems.

A bridging model for parallel computation

The success of the von Neumann model of sequential computation is attributable to the fact that it is an efficient bridge between software and hardware: high-level languages can be efficiently compiled on to this model; yet it can be effeciently implemented in hardware. The author argues that an analogous bridge between software and hardware in required for parallel computation if that is to become as widely used. This article introduces the bulk-synchronous parallel (BSP) model as a candidate for this role, and gives results quantifying its efficiency both in implementing high-level language features and algorithms, as well as in being implemented in hardware.

The complexity of computing the permanent

Abstract It is shown that the permanent function of (0, 1)-matrices is a complete problem for the class of counting problems associated with nondeterministic polynomial time computations. Related counting problems are also considered. The reductions used are characterized by their nontrivial use of arithmetic.

THE COMPLEXITY OF ENUMERATION AND RELIABILITY PROBLEMS

The class of

Random generation of combinatorial structures from a uniform

# P

Universal schemes for parallel communication

-complete problems is a class of computationally eqivalent counting problems (defined by the author in a previous paper) that are at least as difficult as the

Cryptographic limitations on learning Boolean formulae and finite automata

NP

A SCHEME FOR FAST PARALLEL COMMUNICATION

-complete problems. Here we show, for a large number of natural counting problems for which there was no previous indication of intractability, that they belong to this class. The technique used is that of polynomial time reduction with oracles via translations that are of algebraic or arithmetic nature.

Computational limitations on learning from examples

Abstract The class of problems involving the random generation of combinatorial structures from a uniform distribution is considered. Uniform generation problems are, in computational difficulty, intermediate between classical existence and counting problems. It is shown that exactly uniform generation of ‘efficiently verifiable’ combinatorial structures is reducible to approximate counting (and hence, is within the third level of the polynomial hierarchy). Natural combinatorial problems are presented which exhibit complexity gaps between their existence and generation, and between their generation and counting versions. It is further shown that for self-reducible problems, almost uniform generation and randomized approximate counting are inter-reducible, and hence, of similar complexity.

Completeness classes in algebra

In this paper we isolate a combinatorial problem that, we believe, lies at the heart of this question and provide some encouragingly positive solutions to it. We show that there exists an N-processor <underline>realistic</underline> computer that can simulate arbitrary <underline>idealistic</underline> N-processor parallel computations with only a factor of O(log N) loss of runtime efficiency. The main innovation is an O(log N) time randomized routing algorithm. Previous approaches were based on sorting or permutation networks, and implied loss factors of order at least (log N)<supscrpt>2</supscrpt>.