We demonstrate that efficient meta-learning can be achieved via end-to-end training of deep neural networks with memory distributed across layers. The persistent state of this memory assumes the entire burden of guiding task adaptation. Moreover, its distributed nature is instrumental in orchestrating adaptation.

Rule-based adaptation is a foundational approach to self-adaptation, characterized by its human readability and rapid response. However, building high-performance and robust adaptation rules is often a challenge because it essentially involves searching the optimal design in a complex (variables) space. In response, this paper attempt to employ large language models (LLMs) as a optimizer to construct and optimize adaptation rules, leveraging the common sense and reasoning capabilities inherent in LLMs. Preliminary experiments conducted in SWIM have validated the effectiveness and limitation of our method.

We consider neural network models for stochastic nonlinear dynamical systems where measurements of the variable of interest are only avail(cid:173) able at irregular intervals i.e. most realizations are missing. Difficulties arise since the solutions for prediction and maximum likelihood learn(cid:173) ing with missing data lead to complex integrals, which even for simple cases cannot be solved analytically. In this paper we propose a spe(cid:173) cific combination of a nonlinear recurrent neural predictive model and a linear error model which leads to tractable prediction and maximum likelihood adaptation rules. In particular, the recurrent neural network can be trained using the real-time recurrent learning rule and the linear error model can be trained by an EM adaptation rule, implemented us(cid:173) ing forward-backward Kalman filter equations. The model is applied to predict the glucose/insulin metabolism of a diabetic patient where blood glucose measurements are only available a few times a day at irregular intervals.

In this paper, we focus on exploiting neural networks for the analysis and planning stage in self-adaptive architectures. The studied motivating cases in the paper involve existing (legacy) self-adaptive architectures and their adaptation logic, which has been specified by logical rules. We further assume that there is a need to endow these systems with the ability to learn based on examples of inputs and expected outputs. One simple option to address such a need is to replace the reasoning based on logical rules with a neural network. However, this step brings several problems that often create at least a temporary regress. The reason is the logical rules typically represent a large and tested body of domain knowledge, which may be lost if the logical rules are replaced by a neural network. Further, the black-box nature of generic neural networks obfuscates how the systems work inside and consequently introduces more uncertainty. In this paper, we present a method that makes it possible to endow an existing self-adaptive architectures with the ability to learn using neural networks, while preserving domain knowledge existing in the logical rules. We introduce a continuum between the existing rule-based system and a system based on a generic neural network. We show how to navigate in this continuum and create a neural network architecture that naturally embeds the original logical rules and how to gradually scale the learning potential of the network, thus controlling the uncertainty inherent to all soft computing models. We showcase and evaluate the approach on representative excerpts from two larger real-life use cases.

Differential equations have been widely and successfully used to describe the macroscopic or global behaviour of systems in general. With inhomogeneous, time-varying, specific, and often non-linear interactions, the dynamics of complex systems is in contrast more efficiently described by local rules and thus in an algorithmic and local or microscopic manner. In order to create computer models of systems, particularly complex systems, we recently presented a so-called allagmatic method including a system metamodel with two concrete implementations of cellular automata and artificial neural networks. Guidance from philosophy and its general concepts were helpful in these first studies focusing on computer modelling and simulation as well as feasibility. A rigorous mathematical formalism of the system metamodel, however, is still missing. This would not only more precisely describe and define the system metamodel, it would also further generalise it and with that extend its reach to formal treatment in applied mathematics and computational mathematics as well as extend its applicability to other mathematical models such as agent-based models. In this study, a mathematical definition of the system metamodel and its model parameters is provided and the creation of concrete mathematical models, i.e. cellular automata and artificial neural networks, from it is proved.

A self-adaptive system can modify its own structure and behavior at runtime based on its perception of the environment, of itself and of its requirements. To develop a self-adaptive system, software developers codify knowledge about the system and its environment, as well as how adaptation actions impact on the system. However, the codified knowledge may be insufficient due to design time uncertainty, and thus a self-adaptive system may execute adaptation actions that do not have the desired effect. Online learning is an emerging approach to address design time uncertainty by employing machine learning at runtime. Online learning accumulates knowledge at runtime by, for instance, exploring not-yet executed adaptation actions. We address two specific problems with respect to online learning for self-adaptive systems. First, the number of possible adaptation actions can be very large. Existing online learning techniques randomly explore the possible adaptation actions, but this can lead to slow convergence of the learning process. Second, the possible adaptation actions can change as a result of system evolution. Existing online learning techniques are unaware of these changes and thus do not explore new adaptation actions, but explore adaptation actions that are no longer valid. We propose using feature models to give structure to the set of adaptation actions and thereby guide the exploration process during online learning. Experimental results involving four real-world systems suggest that considering the hierarchical structure of feature models may speed up convergence by 7.2% on average. Considering the differences between feature models before and after an evolution step may speed up convergence by 64.6% on average. [...]

Implementation of the case-based reasoner in rules and objects using a commercial knowledge-based system shell is described. Although some refinements remain, the performance of the case-based reasoner has met its design goals. The chemical industry is heavily regulated. Every hazardous chemical product must have a set of shipping descriptions that conform to strict regulations issued by the Department of Transportation (DOT), the International Maritime Organization (IMO), and the International Air Transport Association (IATA). Shipping descriptions provide a concise characterization of the hazards a chemical can present during transportation (figure 1). Failing to comply with transport regulations can result in penalties ranging from delayed shipments to heavy fines or even incarceration of corporate officials. In addition, each chemical product has a material safety data sheet (MSDS) that conforms to Occupational Safety and Health (OSHA) and American National Standards Institute (ANSI) standards. Unlike shipping descriptions, MSDSs are lengthy documents of 8 to 10 pages that provide a detailed description of the health hazards a product can pose in the workplace (figure 2). They also contain information on procedures for storing, handling, and disposing of a chemical. Inadequately prepared MSDSs can lead to substantial product-liability lawsuits against the company if the product is involved in an industrial accident. The ultimate goal of these regulations is to ensure proper communication of health and safety information for the protection of the public. Air Products is committed to the initiative of the Chemical Manufacturers Association (CMA) known as Responsible Care. This initiative focuses on the safe manufacturing, distribution, use, recycling, and disposal of chemicals. Proper communication through accurate shipping descriptions and full disclosure of hazard information in the MSDS plays a key role in fulfilling obligations under Responsible Care. Maintaining shipping descriptions and MSDSs requires a major effort. Most corporate systems are intensely manual.

In case-based reasoning (CBR), problems are solved by retrieving prior cases and adapting their solutions to fit; learning occurs as new cases are stored. Controlling the growth of the case base is a fundamental problem, and research on case-base maintenance has developed methods for compacting case bases while maintaining system competence, primarily by competence-based deletion strategies assuming static case adaptation knowledge. This paper proposes adaptation-guided case-base maintenance (AGCBM), a case-base maintenance approach exploiting the ability to dynamically generate new adaptation knowledge from cases. In AGCBM, case retention decisions are based both on cases' value as base cases for solving problems and on their value for generating new adaptation rules. he paper illustrates the method for numerical prediction tasks (case-based regression) in which adaptation rules are generated automatically using the case difference heuristic. In comparisons of AGCBM to five alternative methods in four domains, for varying case base densities, AGCBM outperformed the alternatives in all domains, with greatest benefit at high compression.

Adaptation has long been considered as the Achilles' heel of case-based reasoning since it requires some domain-specific knowledge that is difficult to acquire. In this paper, two strategies are combined in order to reduce the knowledge engineering cost induced by the adaptation knowledge (AK) acquisition task: AK is learned from the case base by the means of knowledge discovery techniques, and the AK acquisition sessions are opportunistically triggered, i.e., at problem-solving time.

In case-based reasoning, the adaptation of a source case in order to solve the target problem is at the same time crucial and difficult to implement. The reason for this difficulty is that, in general, adaptation strongly depends on domain-dependent knowledge. This fact motivates research on adaptation knowledge acquisition (AKA). This paper presents an approach to AKA based on the principles and techniques of knowledge discovery from databases and data-mining. It is implemented in CABAMAKA, a system that explores the variations within the case base to elicit adaptation knowledge. This system has been successfully tested in an application of case-based reasoning to decision support in the domain of breast cancer treatment.