CHILD: A First Step Towards Continual Learning

Continual learning is the constant development of increasingly complex behaviors; the process of building more complicated skills on top of those already developed. A continual-learning agent should therefore learn incrementally and hierarchically. This paper describes CHILD, an agent capable of Continual, Hierarchical, Incremental Learning and Development. CHILD can quickly solve complicated non-Markovian reinforcement-learning tasks and can then transfer its skills to similar but even more complicated tasks, learning these faster still.     

A First Step Towards Adaptive Control for Linear Systems in Max Algebra

Alinear system theory has been developed for discrete event systemssubject to synchronization. We are interested in the just intime control of such systems in presence of a possible mismatchbetween the system and its model. Taking such mismatch into accountin controller synthesis prevents us from using the usual openloop control structure. The approach we propose to synthesizethe controller is inspired by the principle of the indirect adaptivecontrol well known in conventional system theory. Its designleads to solve an optimal tracking problem under an equalityconstraint. Preliminary results about the estimation of modelsare presented.     

A First Step Towards the Use of Proper General Decomposition Method for Structural Optimization

In structural optimization, the implicit nature of the cost function with respect to the optimization parameters, i.e. through the solution of the structural problem calculated with fixed values of these parameters, leads to prohibitive computations whatever the adopted formulation.     

A History of First Step Fallacies

In the 1960s, without realizing it, AI researchers were hard at work finding the features, rules, and representations needed for turning rationalist philosophy into a research program, and by so doing AI researchers condemned their enterprise to failure. About the same time, a logician, Yehoshua Bar-Hillel, pointed out that AI optimism was based on what he called the ??first step fallacy??. First step thinking has the idea of a successful last step built in. Limited early success, however, is not a valid basis for predicting the ultimate success of one??s project. Climbing a hill should not give one any assurance that if he keeps going he will reach the sky. Perhaps one may have overlooked some serious problem lying ahead. There is, in fact, no reason to think that we are making progress towards AI or, indeed, that AI is even possible, in which case claiming incremental progress towards it would make no sense. In current excited waiting for the singularity, religion and technology converge. Hard headed materialists desperately yearn for a world where our bodies no longer have to grow old and die. They will be transformed into information, like Google digitizes old books, and we will achieve the promise of eternal life. As an existential philosopher, however, I suggest that we may have to overcome the desperate desire to digitalize our bodies so as to achieve immortality, and, instead, face up to and maybe even enjoy our embodied finitude.     

The Grammar of Movement: A Step Towards a Corporeal Architecture

This research uses shape grammars as the basis of a computational tool to explore the relationship between the human body in motion and space, aiming to develop further knowledge about cognition and architecture. Artistic and scientific tools and methods already used to develop these concepts are being studied in order to create a new tool that will help us to understand, through simulation, how the body mediated through architecture can influence human cognitive response and thus behaviour. The goal is a methodology for the design of a “corporeal architecture” that can create a naturally immersive environment in which the ability of its geometry and physical properties to conduct or induce body movements in space for specific purposes can generate experience. Also discussed is the potential of the tool proposed for the study of the human body in movement as a generative strategy in architecture. In describing the parameters and criteria chosen to develop our software, we exemplify briefly how a shape grammar, as a system of rules, can be used to generate sequences of actions, establishing the idea that human behaviour in space can be composed as choreography and provide a means of considering architectural space not only in terms of shape but particularly in terms of life.     

SALA: A Self-Adaptive Learning Algorithm—Towards Efficient Dynamic Route Guidance in Urban Traffic Networks

In order to alleviate traffic congestion for vehicles in urban networks, most of current researches mainly focused on signal optimization models and traffic assignment models, or tried to recognize the interaction between signal control and traffic assignment. However, these methods may not be able to provide fast and accurate route guidance due to the lack of individual traffic demands, real-time traffic data and dynamic cooperation between vehicles. To solve these problems, this paper proposes a dynamic and real-time route selection model in urban traffic networks (DR2SM), which can supply a more accurate and personalized strategy for vehicles in urban traffic networks. Combining the preference for alternative routes with real-time traffic conditions, each vehicle in urban traffic networks updates its route selection before going through each intersection. Based on its historical experiences and estimation about route choices of the other vehicles, each vehicle uses a self-adaptive learning algorithm to play congestion game with each other to reach Nash equilibrium. In the route selection process, each vehicle selects the user-optimal route, which can maximize the utility of each driving vehicle. The results of the experiments on both synthetic and real-world road networks show that compared with non-cooperative route selection algorithms and three state-of-the-art equilibrium algorithms, DR2SM can effectively reduce the average traveling time in the dynamic and uncertain urban traffic networks.     

The Annotated Assistant: A Step Towards Human Engineering

The Hatter's watch nicely illustrates the effect of idiosynracy in system design. Really, a watch could provide any number of features, but most watches designed for people put a high priority on telling the correct time of day. Thus, the Hatter's watch is an excellent example of bad human engineering. By human engineering we mean "the selection among design alternatives so as to relate to people." Carroll's stopped watch is the ultimate in poor human engineering because the user must do all the work.     

Natural-Language Interface for CAD: A First Step

Though the ubiquity of CAD tools in state-of-the-art VLSI design has forced the issue of user interfaces for CAD to prominence, one type of user interface that has not been investigated hitherto is the naturallanquage interface. The CAD domain is substantially more complex than the domains for which previous natural-language interfaces have been built. This implies that a natural-language interface for CAD cannot be based on previous approaches to language understanding; a new approach is required. Our approach emphasizes flexibility, parallelism, and redundancy, and is embodied in an interface called Cleopatra. Ultimately, Cleopatra will be an interface for an ?integrated design environment.? As a first step towards that goal, but also as a useful tool in its own right, Cleopatra currently deals with circuit-simulation post-processing.     

Pulmonary Diffusion—A Model Study

Compared to even the most complex engineering systems, physiological systems are enormously difficult ones in which to conduct direct parametric studies. In this paper, we look at the application of computer simulation techniques to the study of pulmonary diffusion—a physiological process that is especially difficult to study in vivo ( i.e., experiments conducted on a living subject).     

Muse—A Computer Assisted Verification System

Muse is a verification system which extends the collection of tools developed by SRI International for their Hierarchical Development Methodology (HDM). It enhances the SRI system by providing a capability for proving invariants and constraints for the state machine described by a specification written in SPECIAL (the specification language of HDM). In particular, it enables one to use the HDM system to meet the requirements for formal verification in a National Computer Security Center A1 evaluation of a secure operating system. In addition to the tools provided by SRI, Muse has a parser, a facility to handle multiple modules, a formula generator, and a theorem prover. The theorem prover has a number of interesting features designed to facilitate human direction of the proving process. In concept, it is open-ended. We introduce the notion of a theorem prover kernel as a device for ensuring the logical soundness of the prover in the face of continual improvements to its functionality.     

SEES—A Software testing Environment Support System

SEES is a database system to support program testing. The program database is automatically created during the compilation of the program by a compiler built using the YACC compiler-compiler.     

*MOD—A Language for Distributed Programming

Distributed programming is characterized by high communications costs and the inability to use shared variables and procedures for interprocessor synchronization and communication. *MOD is a high-level language system which attempts to address these problems by creating an environment conducive to efficient and reliable network software construction. Several of the *MOD distributed programming constructs are discussed as well as an interprocessor communication methodology. Examples illustrating these concepts are drawn from the areas of network communication and distributed process synchronization.     

Scheme-79 — Lisp on a Chip

The Scheme-79 single-chip microcomputer implements an automatic storage allocation system and an interrupt facility to support direct interpretation of a variant of the Scheme dialect of Lisp.     

ESPRESO—A System for Process Control Software Specification

This paper outlines a specification system for process control software, named ESPRESO, which was developed at the Nuclear Research Center, Karlsruhe, West Germany. ESPRESO is based on some new ideas, which are combined with elements taken from other systems. ESPRESO consists of a set of concepts, a specification language, a tool for the management, evaluation and validation of specifications, and the method how to use the system. Language, tool, and method are carefully adapted to the concepts. The primary aim was to demonstrate some features of a specification system which are currently not available, rather than to provide a new tool for the software market.     

DISSECT—A Symbolic Evaluation and Program Testing System

The basic features of the DISSECT symbolic testing tool are described. Usage procedures are outlined and the special advantages of the tool are summarized. Cost estimates for using the tool are provided and the results of experiments to determine its effectiveness are included. The background and history of the development of the tool are outlined. The availability of the tool is described and a listing of reference materials is included.     

Higher Order Software—A Methodology for Defining Software

The key to software reliability is to design, develop, and manage software with a formalized methodology which can be used by computer scientists and applications engineers to describe and communicate interfaces between systems. These interfaces include: software to software; software to other systems; software to management; as well as discipline to discipline within the complete software development process. The formal methodology of Higher Order Software (HOS), specifically aimed toward large-scale multiprogrammed/multiprocessor systems, is dedicated to systems reliability. With six axioms as the basis, a given system and all of its interfaces is defined as if it were one complete and consistent computable system. Some of the derived theorems provide for: reconfiguration of real-time multiprogrammed processes, communication between functions, and prevention of data and timing conflicts.     

LM2— A Logic Machine Minicomputer

In 1963 Clark and Molnar¹developed the LINC computer (Laboratory Instrument Computer), principally for use in biomedical research. Several of the features of this machine (autoindexing, LINC tapes, 12-bit word, data break) later appeared in the PDP-5 and, still later, in the PDP-8. It is probably fair to give Clark and Molnar credit for thus starting the minicomputer revolution.     

Software Reliability—Status and Perspectives

It is essential to assess the reliability of digital computer systems used for critical real-time control applications (e.g., nuclear power plant safety control systems). This involves the assessment of the design correctness of the combined hardware/software system as well as the reliability of the hardware. In this paper we survey methods of determining the design correctness of systems as applied to computer programs.     

Generalized Performance of Concatenated Quantum Codes—A Dynamical Systems Approach

We apply a dynamical systems approach to concatenation of quantum error correcting codes, extending and generalizing the results of Rahn to both diagonal and nondiagonal channels. Our point of view is global: instead of focusing on particular types of noise channels, we study the geometry of the coding map as a discrete-time dynamical system on the entire space of noise channels. In the case of diagonal channels, we show that any code with distance at least three corrects (in the infinite concatenation limit) an open set of errors. For Calderbank–Shor–Steane (CSS) codes, we give a more precise characterization of that set. We show how to incorporate noise in the gates, thus completing the framework. We derive some general bounds for noise channels, which allows us to analyze several codes in detail.