Class (philosophy)

About: Class (philosophy) is a research topic. Over the lifetime, 821 publications have been published within this topic receiving 28000 citations.

Name Collision in Multiple Classification Hierarchies

Abstract: Supporting multiple classification in object-oriented programming languages is the topic of discussion in this paper. Supporting multiple classification gives rise to one important question --- namely the question of inheritance of attributes with identical names from multiple paths in the classification hierarchy. The problem is to decide how these multiple classification paths are reflected in the class being defined. One of the conclusions in this paper is, that by choosing strict and simple inheritance rules, one is excluding some particular usages of multiple classification. This leads to the notion of attribute-resolution at class definition, which means that the programmer in some cases is forced or allowed to resolve the potential ambiguity of the inherited names. The concept of attribute-resolution is managed through the identification of two conceptually different utilizations of specialization (unification and intersection), and two different attribute properties (plural and singleton) to guide the attribute-resolution.

Constrained Few-shot Class-incremental Learning

Abstract: Continually learning new classes from fresh data without forgetting previous knowledge of old classes is a very challenging research problem. Moreover, it is imperative that such learning must respect certain memory and computational constraints such as (i) training samples are limited to only a few per class, (ii) the computational cost of learning a novel class remains constant, and (iii) the memory footprint of the model grows at most linearly with the number of classes observed. To meet the above constraints, we propose C-FSCIL, which is architecturally composed of a frozen meta-learned feature extractor, a trainable fixed-size fully connected layer, and a rewritable dynamically growing memory that stores as many vectors as the number of encountered classes. C-FSCIL provides three update modes that offer a trade-off between accuracy and compute-memory cost of learning novel classes. C-FSCIL exploits hyperdimensional embedding that allows to continually express many more classes than the fixed dimensions in the vector space, with minimal interference. The quality of class vector representations is further improved by aligning them quasi-orthogonally to each other by means of novel loss functions. Experiments on the CIFAR100, mini-ImageNet, and Omniglot datasets show that C-FSCIL outperforms the baselines with remarkable accuracy and compression. It also scales up to the largest problem size ever tried in this few-shot setting by learning 423 novel classes on top of 1200 base classes with less than 1.6% accuracy drop. Our code is available at https://github.com/IBM/constrained-FSCIL.

Extracting frame conditions from operation contracts

Abstract: In behavioral modeling, operation contracts defined by pre- and postconditions describe the effects on model properties (i.e., model elements such as attributes, links, etc.) that are enforced by an operation. However, it is usually omitted which model properties should not be modified. Defining so-called frame conditions can fill this gap. But, thus far, these have to be defined manually - a time-consuming task. In this work, we propose a methodology which aims to support the modeler in the definition of the frame conditions by extracting suggestions based on an automatic analysis of operation contracts provided in OCL. More precisely, the proposed approach performs a structural analysis of pre- and postconditions together with invariants in order to categorize which class and object properties are clearly "variable" or "unaffected" - and which are "ambiguous", i.e. indeed require a more thorough inspection. The developed concepts are implemented as a prototype and evaluated by means of several example models known from the literature.

Mean Field Games with monotonous interactions through the law of states and controls of the agents

Abstract: We consider a class of Mean Field Games in which the agents may interact through the statistical distribution of their states and controls. It is supposed that the Hamiltonian behaves like a power of its arguments as they tend to infinity, with an exponent larger than one. A monotonicity assumption is also made. Existence and uniqueness are proved using a priori estimates which stem from the monotonicity assumptions and Leray-Schauder theorem. Applications of the results are given.

Weighted Fractional Calculus: A General Class of Operators

Abstract: We conduct a formal study of a particular class of fractional operators, namely weighted fractional calculus, and its extension to the more general class known as weighted fractional calculus with respect to functions. We emphasise the importance of the conjugation relationships with the classical Riemann–Liouville fractional calculus, and use them to prove many fundamental properties of these operators. As examples, we consider special cases such as tempered, Hadamard-type, and Erdélyi–Kober operators. We also define appropriate modifications of the Laplace transform and convolution operations, and solve some ordinary differential equations in the setting of these general classes of operators.