Andrew Coward

Andrew Coward
Australian National University | ANU · Research School of Computer Science

About

60
Publications
3,626
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
391
Citations
Additional affiliations
January 2004 - present
Australian National University
Position
  • Professor (Associate)

Publications

Publications (60)
Book
This book shows how to understand higher cognition in terms of brain anatomy, physiology and chemistry. Natural selection pressures have resulted in all information processes in the brain being one of just two general types: condition definition/detections and behavioural recommendation definition/integrations. Using these information process types...
Chapter
Full-text available
This chapter is the introduction to the book Brain Mechanisms: Linking Cognitive Phenomena to Neuron Activity
Chapter
Full-text available
In computers there are two key architectures: the functional architecture, which is critical for understanding how to use a computer, and the system architecture, which is critical for understanding the design of the computer. The functional architecture divides up the system into user feature components called applications. The system architecture...
Article
Full-text available
Theoretical arguments demonstrate that practical considerations, including the needs to limit physiological resources and to learn without interference with prior learning, severely constrain the anatomical architecture of the brain. These arguments identify the hippocampal system as the change manager for the cortex, with the role of selecting the...
Article
Full-text available
The brain uses computational primitives that are analogous with but qualitatively different from the computational primitives used in electronic computer systems. The primary computational primitives of the brain are described, and their implementation in anatomy and physiology discussed. Combinations and sequences of these primitives implement cog...
Chapter
The environment in which a brain operates contains many different objects. Some objects urgently require a behavioural response, for others a response is less urgent or not required. Developing a response to an object requires detection of receptive fields that can discriminate between different types of object (≈ objects) within the sensory inform...
Chapter
We have established in Chaps. 9 and 10 that causal descriptions of relatively simple cognitive phenomena at high-level using the condition definition/detection and behavioural recommendation information models can be mapped into more detailed causal descriptions at anatomical, physiological and neurochemical levels. Hence we can have confidence tha...
Chapter
There are many different types of neuron organized into many different anatomical structures in the mammal brain. Many of these structures are connected together, sometimes by multiple routes, and often indirectly via other structures. Neuron physiology involves large numbers of different chemicals interacting by many complex pathways.
Chapter
Our brains play a role in almost everything we do, but in this book the focus is on activities under the general label of higher cognition, including memory and learning, speech, emotions, intellectual disciplines, planning, consciousness and self awareness.
Chapter
As discussed in Chap. 7, a number of practical requirements tend to constrain the physical resources of a complex learning system into specific forms called the recommendation architecture. These forms include the existence of hierarchies of modules and components, with modules and components on different levels using the same two types of informat...
Chapter
In the previous chapter we looked at the phenomena of higher human cognition. The brain is the organ where much of the activity supporting higher cognition takes place, and in this chapter we will provide an overview of major brain anatomy structures, and a simplified overview of neurons and their physiology within those structures. In Chaps. 4 and...
Chapter
A number of practical requirements constrain the physical architecture of any system which must learn to perform a complex combination of behaviours with resources that are not unlimited [932]. In this chapter we will review the nature of these constraints and the way in which they arise. Analogous practical requirements constrain the architecture...
Chapter
A neuron receives large numbers of input stimuli from other neurons and generates an output action potential in response to those inputs. In order to preserve past learning the integration algorithm that determines outputs from inputs must be relatively stable, but to implement short term or longer term learning it must be changeable on various tim...
Chapter
This book addresses the issue of how higher cognition can be understood in terms of the neuroanatomy, neurophysiology, and neurochemistry of the human brain. Much is known about the chemical processes within and between neurons, and this knowledge has, for example, led to practical applications in the area of drugs to treat brain diseases and defic...
Chapter
In this chapter we will describe the morphology and physiology of neurons in much more detail, with particular emphasis on the ways in which molecules spanning the neuron membrane result in detection and integration of signals received from other neurons. At each point in time a neuron receives large numbers of external signals, most derived from o...
Chapter
What can we expect from a theoretical neuroscience? Experimental work on brains has been carried out on many levels of detail, including behavioural psychology, major anatomical structures, detailed anatomical structures, neurons, and detailed chemical processes within neurons. However, for human understanding of how cognition results from activity...
Chapter
Memory involves the recording of information about experiences, and accessing that information to guide behaviour during later experiences that have some elements of similarity with earlier experiences. One new experience may have many different, potentially relevant elements of similarity with many past experiences.
Book
The book explains how to understand cognition in terms of brain anatomy, physiology and chemistry, using an approach adapted from techniques for understanding complex electronic systems. These techniques create hierarchies of information process based descriptions on different levels of detail, where higher levels contain less information and can t...
Chapter
Full-text available
Natural selection pressures have resulted in the physical resources of the brain being organized into modules that perform different general types of information processes. Each module is made up of submodules performing different information processes of the general type, and each submodule is made up of yet more detailed modules. At the highest l...
Conference Paper
Full-text available
The brain carries out cognitive learning and processing by performing combinations of different types of information processes. Types of information processes are performed by different anatomical structures and implemented in physiology. The information processes performed by different major anatomical structures including the cortex, basal gangli...
Article
A system which performs a complex combination of behaviours has two superficially independent architectures. One is the functional architecture, which separates the behavioural features of the system into feature modules made up of groups of similar behaviours, and defines the interactions between features. The other is the system architecture (alt...
Article
A model is described in which the hippocampal system functions as resource manager for the neocortex. This model is developed from an architectural concept for the brain as a whole within which the receptive fields of neocortical columns can gradually expand but with some limited exceptions tend not to contract. The definition process for receptive...
Conference Paper
A model is described in which the hippocampal system receives inputs from cortical columns throughout the neocortex, uses these inputs to determine the most appropriate columns for declarative information recording in response to the current sensory experience, and generates outputs that drive that recording in the selected columns. Evidence in sup...
Article
Full-text available
A machine with human-like consciousness would be an extremely complex system. Prior work has demonstrated that the way in which information handling resources are organized (the resource architecture) in an extremely complex learning system is constrained within some specific bounds if the available resources are limited, and that there is evidence...
Article
There has been much discussion on what a scientific theory of consciousness would look like, and even whether such a theory is possible. Some common misunderstandings of the nature of theories (e.g., in the physical sciences) have confused the discussion of theories concerning consciousness. Theories in the physical sciences establish hierarchies o...
Chapter
Full-text available
A detailed connectionist architecture is described which is capable of relating psychological behavior to the functioning of neurons and neurochemicals. The need to be able to build, repair and modify current electronic systems with billions of hardware components has been met through a seldom appreciated aspect of the von Neumann architecture: the...
Conference Paper
Full-text available
Any system which must learn to perform a large number of behavioral features with limited information handling resources will tend to be constrained within a set of architectural bounds. Unless design for a system with human like intelligence is performed within these bounds, the system will require excessive resources, and learning will introduce...
Article
The article first addresses the importance of cognitive modeling, in terms of its value to cognitive science (as well as other social and behavioral sciences). In particular, it emphasizes the use of cognitive architectures in this undertaking. Based on this approach, the article addresses, in detail, the idea of a multi-level approach that ranges...
Article
Full-text available
A qualitative explanation for experimental observations of episodic, semantic and procedural memory and the differences between them is provided based on the different ways in which information derived from the senses is recorded and accessed in the recommendation architecture cognitive model.
Article
In the recommendation architecture cognitive model, information can only be recorded in two ways: by permanent recording of information conditions or by a change to the recommendation weight of a condition in favour of a behaviour. Information can only be accessed in four ways: activation of conditions which occur within sensory inputs; activation...
Conference Paper
Full-text available
Categorization is a primary application for conceptual knowledge stored in the human brain. Categorization is often based on similarity, involving apparent use of both prototypes and stored exemplars. Some human categorization appears to be rule or theory based rather than based on similarity. Attempts to model categorization often involve multiple...
Article
Full-text available
Categorization is a primary application for conceptual knowledge stored in the human brain. Categorization is often based on similarity, involving apparent use of both prototypes and stored exemplars. Some human categorization appears to be rule or theory based rather than based on similarity. Attempts to model categorization often involve multiple...
Book
This book is the integrated presentation of a large body of work on understanding the operation of biological brains as systems. The work has been carried out by the author over the last 22 years, and leads to a claim that it is relatively straightforward to understand how human cognition results from and is supported by physiological processes in...
Conference Paper
Full-text available
A device algorithm which permanently records an expanding portfolio of similar conditions is described, along with an architecture in which this algorithm is used to avoid interference between prior and later learning.
Conference Paper
Full-text available
The reasons why machine learning appears limited to the relatively simple control problems are analyzed. A primary issue is that, any condition detected by a learning system acquires multiple behavioural meanings. As the learning continues, the need to preserve these meanings severely constrains the architectural form of the system. A hybrid archit...
Article
In the physical sciences a rigorous theory is a hierarchy of descriptions in which causal relationships between many general types of entity at a phenomenological level can be derived from causal relationships between smaller numbers of simpler entities at more detailed levels. The hierarchy of descriptions resembles the modular hierarchy created i...
Article
Full-text available
Cities can be viewed as information architecture systems. Here, ‘architecture’ is used in the sense of computer architecture – it refers not to the design of buildings, but to how the components of a complex system interact. Information exchange includes the movement of people and goods, personal contact and interactions, telecommunications, as wel...
Article
In the physical sciences a rigorous theory is a hierarchy of descriptions in which causal relationships between many general types of entity at a phenomenological level can be derived from causal relationships between smaller numbers of simpler entities at more detailed levels. The hierarchy of descriptions resembles the modular hierarchy created i...
Article
functional changes without side effects, and permit identification and repair of failure conditions. Analogous needs exist for biological brains. These needs force any system which performs a sufficiently complex combination of functions into a modular hierarchy defined by the requirement that modules on each level are roughly equal and information...
Article
Full-text available
The recommendation architecture is a functional architecture with the capability to heuristically define a complex combination of functions which has been proposed as the architecture adopted by biological brains. This paper reports simulations of a system with that architecture which can heuristically define repetition conditions in a constantly v...
Article
The recommendation architecture has been proposed as a system architecture which can enable a system to learn to perform a complex combination of interrelated functions. The capability of a system with the recommendation architecture to learn to manage complex telecommunication backbone networks has been investigated. A network model with a number...
Article
Full-text available
A fundamental approach of cognitive science is to understand cognitive systems by separating them into modules. Theoretical reasons are described which force any system which learns to perform a complex combination of real time functions into a modular architecture. Constraints on the way modules divide up functionality are also described. The arch...
Chapter
A definition of the scope of a scientific theory of human cognition is proposed in which for any psychological state a corresponding physiological state can be identified, and causal relationships between psychological states have corresponding causal relationships between physiological states. The vital role of a simple functional architecture in...
Article
Full-text available
The technology for the design of systems to perform extremely complex combinations of real-time functionality has developed over a long period. This technology is based on the use of a hardware architecture with a physical separation into memory and processing, and a software architecture which divides functionality into a disciplined hierarchy of...
Article
The nature of a scientific theory of consciousness is defined by comparison with scientific theories in the physical sciences. The differences between physical, algorithmic and functional complexity are highlighted, and the architecture of a functionally complex electronic system created to relate system operations to device operations is compared...
Conference Paper
Full-text available
Functionally complex electronic systems are organized into functional components exchanging unambiguous information. The requirement to exchange unambiguous information results in difficulties in implementing parallel processing and extreme difficulty in implementing any capability to heuristically change functionality based on experience. The reco...
Article
Full-text available
Severe constraints apply to the architecture of any system which uses large numbers of components to deliver complex functionality. If not taken into account, these constraints can invalidate conclusions drawn from simulation of cognitive subsystems. The pattern extraction hierarchy is a connectionist architecture which can satisfy these constraint...
Conference Paper
This paper describes the architecture and circuit techniques used for a neural network compute engine. The neural network application is for general pattern recognition and the architecture is based on an adaptive connection network with majority gates acting as neurons. The paper discusses the training/recall algorithms associated with the network...
Conference Paper
The authors present a new architecture for machine perception of objects using a hierarchical pattern extraction technique. The resulting architecture is a neural network with ordinary logic gates as the neurons and simple heuristic pattern association techniques as the training algorithm. The architecture consists of a multilayer network of neuron...
Book
The book develops a system model for the human brain based on a new physiologically based theory of learning and memory. Arguing that the brain can only be fully understood by application of consistent concepts to both psychology and physiology, he proposes that the necessary concepts for constructing a model of brain function are conditions and be...
Article
The architectural constraints on systems which perform complex combinations of real time functions are described, and the conclusion reached that only a system with the recommendation architecture can learn to perform functionality in such a way that it can add new functionality without disruption to previously learned functionality. The design of...
Article
A model for human cognitive processing is described. The advantages of the model are that it is able to learn complex combinations of capabilities with limited information recording and processing resources; it can bootstrap its memory and cognitive capabilities from experience with very limited, genetically plausible a priori guidance; and modules...
Article
Full-text available
The human brain is a functionally complex system. It can be demonstrated both on theoretical grounds and from design experience with functionally complex electronic systems that practical considerations such as the needs to limit information recording and processing resources and to be able to make functional changes without excessive undesirable s...
Article
Full-text available
A model for cortex information processing is described which depends upon a combination of population, rate and temporal coding for action potential spikes. In this model, binding of information derived from one visual attention object occurs because attention causes a slight (~ 1 millisecond) shift in each spike in the sensory inputs derived from...
Article
Full-text available
There are a set of theoretical bounds on any system which must learn to perform a complex combination of features with limited information handling resources. A cognitive architecture implemented within these bounds has the capability to generate human-like "conscious" behaviours with mechanisms that are physiologically plausible. The Characteristi...
Article
Full-text available
Any architecture for modeling cognitive development must have several general characteristics. It must be possible to learn complex combinations of interacting cognitive capabilities using information derived from the same experience stream. Learning must be bootstrapped from experience with minimal a priori guidance and limited external guidance d...
Article
Full-text available
A system which uses large numbers of devices to perform a complex functionality is forced to adopt a simple functional architecture by the needs to construct copies of, repair, and modify the system. A simple functional architecture means that functionality is partitioned into relatively equal sized components on many levels of detail down to devic...