Open Access
A study of OWA operators learned in convolutional neural networks
(MDPI, 2021) Domínguez Catena, Iris; Paternain Dallo, Daniel; Galar Idoate, Mikel; Institute of Smart Cities - ISC; Universidad Pública de Navarra / Nafarroako Unibertsitate Publikoa
Ordered Weighted Averaging (OWA) operators have been integrated in Convolutional Neural Networks (CNNs) for image classification through the OWA layer. This layer lets the CNN integrate global information about the image in the early stages, where most CNN architectures only allow for the exploitation of local information. As a side effect of this integration, the OWA layer becomes a practical method for the determination of OWA operator weights, which is usually a difficult task that complicates the integration of these operators in other fields. In this paper, we explore the weights learned for the OWA operators inside the OWA layer, characterizing them through their basic properties of orness and dispersion. We also compare them to some families of OWA operators, namely the Binomial OWA operator, the Stancu OWA operator and the expo-nential RIM OWA operator, finding examples that are currently impossible to generalize through these parameterizations.
Open Access
Orness measurements for lattice m-dimensional interval-valued OWA operators
(Elsevier, 2018) Miguel Turullols, Laura de; Paternain Dallo, Daniel; Lizasoain Iriso, María Inmaculada; Ochoa Lezaun, Gustavo; Bustince Sola, Humberto; Estatistika, Informatika eta Matematika; Institute for Advanced Materials and Mathematics - INAMAT2; Institute of Smart Cities - ISC; Estadística, Informática y Matemáticas
Ordered weighted average (OWA) operators are commonly used to aggregate information in multiple situations, such as decision making problems or image processing tasks. The great variety of weights that can be chosen to determinate an OWA operator provides a broad family of aggegating functions, which obviously give diferent results in the aggregation of the same set of data. In this paper, some possible classifications of OWA operators are suggested when they are de ned on m-dimensional intervals taking values on a complete lattice satisfying certain local conditions. A first classification is obtained by means of a quantitative orness measure that gives the proximity of each OWA to the OR operator. In the case in which the lattice is finite, another classification is obtained by means of a qualitative orness measure. In the present paper, several theoretical results are obtained in order to perform this qualitative value for each OWA operator.
Open Access
OWA operators based on admissible permutations
(IEEE, 2019) Paternain Dallo, Daniel; Jin, LeSheng; Mesiar, Radko; Vavríková, Lucia; Bustince Sola, Humberto; Estadística, Informática y Matemáticas; Estatistika, Informatika eta Matematika; Universidad Pública de Navarra / Nafarroako Unibertsitate Publikoa, PJUPNA13
In this work we propose a new OWA operator defined on bounded convex posets of a vector-lattice. In order to overcome the non-existence of a total order, which is necessary to obtain a non-decreasing arrangement of the input data, we use the concept of admissible permutation. Based on it, our proposal calculates the different ways in which the input vector could be arranged, always respecting the partial order. For each admissible arrangement, we calculate an intermediate value which is finally collected and averaged by means of the arithmetic mean. We analyze several properties of this operator and we give some counterexamples of those properties of aggregation functions which are not satisfied.
Open Access
Additional feature layers from ordered aggregations for deep neural networks
(IEEE, 2020) Domínguez Catena, Iris; Paternain Dallo, Daniel; Galar Idoate, Mikel; Institute of Smart Cities - ISC; Universidad Pública de Navarra / Nafarroako Unibertsitate Publikoa
In the last years we have seen huge advancements in the area of Machine Learning, specially with the use of Deep Neural Networks. One of the most relevant examples is in image classification, where convolutional neural networks have shown to be a vital tool, hard to replace with any other techniques. Although aggregation functions, such as OWA operators, have been previously used on top of neural networks, usually to aggregate the outputs of different networks or systems (ensembles), in this paper we propose and explore a new way of using OWA aggregations in deep learning. We implement OWA aggregations as a new layer inside a convolutional neural network. These layers are used to learn additional order-based information from the feature maps of a certain layer, and then the newly generated information is used as a complement input for the following layers. We carry out several tests introducing the new layer in a VGG13-based reference network and show that this layer introduces new knowledge into the network without substantially increasing training times.
Embargo
Some characterizations of lattice OWA operators
(World Scientific Publishing Company, 2017) Miguel Turullols, Laura de; Paternain Dallo, Daniel; Lizasoain Iriso, María Inmaculada; Ochoa Lezaun, Gustavo; Bustince Sola, Humberto; Automática y Computación; Automatika eta Konputazioa; Matemáticas; Matematika
Ordered Weighted Averaging (OWA) operators are a family of aggregation which fusion data. If the data are real numbers, then OWA operators can be characterized either as an special kind of Choquet integral or simply as an arithmetic mean of the given values previously ordered. This paper analyzes the possible generalizations of these characterizations when OWA operators are de ned on a complete lattice. In addition, the set of all n -ary OWA operators is studied as a sublattice of the lattice of all the n -ary aggregation functions de ned on a distributive lattice.
Open Access
Orness for real m-dimensional interval-valued OWA operators and its application to determine a good partition
(Taylor & Francis, 2019) Miguel Turullols, Laura de; Paternain Dallo, Daniel; Lizasoain Iriso, María Inmaculada; Ochoa Lezaun, Gustavo; Bustince Sola, Humberto; Estatistika, Informatika eta Matematika; Institute of Smart Cities - ISC; Institute for Advanced Materials and Mathematics - INAMAT2; Estadística, Informática y Matemáticas; Universidad Pública de Navarra / Nafarroako Unibertsitate Publikoa, PJUPNA1
Ordered Weighted Averaging (OWA) operators are a profusely applied class of averaging aggregation functions, i.e. operators that always yield a value between the minimum and the maximum of the inputs. The orness measure was introduced to classify the behavior of the OWA operators depending on the weight vectors. Defining a suitable orness measure is an arduous task when we deal with OWA operators defined over more intricate spaces, such us intervals or lattices. In this work we propose a suitable definition for the orness measure to classify OWA operators defined on the set of m-dimensional intervals taking real values in [0, 1]. The orness measure is applied to decide which is the best partition of a continuous range that should be divided into four linguistic labels. This example shows the good behavior of the proposed orness measure.