Deterministic bridge regression for compressive classification
Pattern classification using a compact representation is a crucial component of machine intelligence. Specifically, it is essential to learn a model with well-regulated parameters to achieve good generalization. Bridge regression provides a ...
Highlights
- An analytic solution for bridge regression has been derived for overdetermined and under-determined systems.
- The solution in dual form can be useful for few-shots learning when data is scarce.
- The analytic solution in primal and ...
Cluster-based data relabelling for classification
Linear classifiers are generally simpler and more explainable than their nonlinear variants. They can achieve satisfactory classification performance on linearly separable data, but not on nonlinear data. So, linear classifiers need extending, ...
A randomized network approach to multifractal texture descriptors
Texture recognition is one of the most important tasks in computer vision, with numerous applications in several areas. Despite the recent success of end-to-end deep learning models in image recognition in general, when it comes to texture images,...
Highlights
- A randomized neural network applied over the space of a multifractal transform is proposed for texture recognition.
- A multiscale approach is developed to provide more robust descriptors.
- The performance of the proposed descriptors ...
CasTformer: A novel cascade transformer towards predicting information diffusion
Predicting information diffusion cascade is an essential task in social networks. We mainly focus on predicting the size of the information cascade. The relationships inside a cascade are diverse, including global and relative spatio-temporal ...
Highlights
- Our proposed CasTformer can fully capture the diverse cascade relationships.
- The STPE effectively marks the global spatio-temporal position information.
- The RRB complements the complex local interaction information.
- The self-...
Graph-coupled time interval network for sequential recommendation
Modeling the dynamics of sequential patterns (i.e., sequential recommendation) has obtained great attention, where the key problem is how to infer the next interesting item according to users' historical actions. Owing to high efficiency and ...
Highlights
- Proposing a graph-coupled time interval network for sequential recommendation.
- Designing a category-aware graph propagation module to better learn user and item embeddings.
- Devising a time-aware self-attention mechanism to capture ...
Switching adaptive event-triggered consensus control for MASs subject to sequential scaling attacks and transmission delays
The secure consensus issue has been examined for multi-agent systems that are vulnerable to sequential scaling attacks with network communication delays in this paper. It is suggested to use the innovative sampled data based switching event-...
Multi-label feature selection based on stable label relevance and label-specific features
Multi-label feature selection can efficiently handle large amounts of multi-label data. However, two pressing issues remain in sparse learning for multi-label data. First, many methods explore label relevance through the original label matrix, ...
Scores of hesitant fuzzy elements revisited: “Was sind und was sollen”
This paper revolves around the notion of score for hesitant fuzzy elements, the constituent parts of hesitant fuzzy sets. Scores allow us to reduce the level of uncertainty of hesitant fuzzy sets to classical fuzzy sets, or to rank alternatives ...
Highlights
- We revise the definition and properties of scores for hesitant sets.
- New scores are introduced for some classes of infinite subsets of the unit interval.
- A score avoiding ties on closed non-degenerated intervals is built as a by-...
An ordered feature recognition method based on ranking separability
Ordered and unordered features generally coexist in real-world ordinal classification tasks. For this kind of task, the classification performance can be improved by distinguishing between ordered and unordered features, so it is necessary to ...
Highlights
- Ranking separability is proposed, and it can measure the distance between classes from the perspective of features.
- This method considers comprehensively the relationships between all features and class, and can recognize ordered ...
Accelerating wargaming reinforcement learning by dynamic multi-demonstrator ensemble
Deep Reinforcement Learning (DRL) has become a promising technique to deal with tough wargaming decision-making problems. However, DRL suffers an inherent problem of low learning efficiency and it often requires massive cost of training steps, ...
Consensus of three-way group decision with weight updating based on a novel linguistic intuitionistic fuzzy similarity
Three-way group decision-making is an important direction of study in the fields of three-way decision-making and granular computing and has received much attention from scholars in recent years. According to previous studies, three-way group-...
On the approximation of Euclidean SL via geometric method
Single-linkage with distance-r stopping condition (SL) is a classical clustering technique, which can discover arbitrary shaped clusters. However, its O ( n 2 ) worst-case time complexity is still a challenge, where n is the number of points in ...
K-filter-based distributed adaptive output cluster consensus under cooperative-competitive networks
This paper investigates the K-filter-based output cluster consensus problem under cooperative-competitive networks. The agents in the network are divided into several clusters where the relationships among agents belonging to the same cluster are ...
An efficient randomized QLP algorithm for approximating the singular value decomposition
The rank-revealing pivoted QLP decomposition approximates the computationally prohibitive singular value decomposition (SVD) via two consecutive column-pivoted QR (CPQR) decomposition. It furnishes information on all four fundamental subspaces of ...
Reward shaping using convolutional neural network
In this paper, we propose Value Iteration Network for Reward Shaping (VIN-RS), a potential-based reward shaping mechanism using Convolutional Neural Network (CNN). The proposed VIN-RS embeds a CNN trained on computed labels using the message ...
Stabilization of DFIG-based wind turbine with active and reactive power: A coupling memory state-feedback control scheme
In this paper, we examine a stabilization problem for the active and reactive power of doubly fed induction generator (DFIG)-based wind turbine (WT) through fuzzy coupling memory state-feedback control (CMSFC). In order to narrate the nonlinear ...
Connecting concept lattices with bonds induced by external information
- Ondrej Krídlo,
- Domingo López-Rodríguez,
- Lubomir Antoni,
- Peter Eliaš,
- Stanislav Krajči,
- Manuel Ojeda-Aciego
In Formal Concept Analysis (FCA), L-bonds represent relationships between L-formal contexts. Choosing the appropriate bond between L-fuzzy formal contexts is an important challenge for its application in recommendation tasks. Recent work ...
Multi-robot task allocation methods: A fuzzy optimization approach
Response-threshold methods stand out among the different developed swarm-like methodologies that address the task allocation problem, which must be faced in multi-robot systems in order to assign to each robot the best task to perform at each ...
Highlights
- A task-allocation strategy for RTM methods dealing with deadlines is proposed.
- The robot stimuli are modeled using fuzzy sets and aggregation functions.
- Robots decide the best task to go through a fuzzy optimization technique.
- ...
A study of dynamic populations in geometric semantic genetic programming
Allowing the population size to variate during the evolution can bring advantages to evolutionary algorithms (EAs), retaining computational effort during the evolution process. Dynamic populations use computational resources wisely in several ...
Distributed resilient fusion estimation for resource-limited CPSs under hybrid attacks
This article investigates the problem of distributed fusion estimation for cyber-physical systems with resource-constraints in the presence of hybrid attacks. First, an event-based protector is designed by resorting to the Gaussian mixture model (...
Decomposition of idempotent pseudo-uninorms via ordinal sum
The decomposition of idempotent pseudo-uninorms is investigated. We show that each idempotent pseudo-uninorm on the unit interval can be decomposed into an ordinal sum of trivial semigroups and non-commutative idempotent semigroups defined on two ...
Time series compression based on reinforcement learning
Nowadays, sensors and signal catchers in various fields are capturing time-series data all the time, and time-series data are exploding. Due to the large storage space requirements and redundancy, many compression techniques for time series have ...
Joint graph entropy knowledge distillation for point cloud classification and robustness against corruptions
Classification tasks in 3D point clouds often assume that class events are independent and identically distributed (IID), although this assumption destroys the correlation between classes. This study proposes a classification strategy, Joint G...
Highlights
- Re-examine the point cloud task from the perspective of non-IID.
- Constructing class-to-class relationship graphs through joint probabilities.
- Realize inter-class domain transfer through knowledge distillation.
- Knowledge ...
Multiple structured latent double dictionary pair learning for cross-domain industrial process monitoring
Process data collected from real-world industrial operating environments have different distributions and lack real-time labeled samples, which causes the performance of process monitoring to decline. In this paper, a multiple structured latent ...
Mitigating the performance sacrifice in DP-satisfied federated settings through graph contrastive learning
Currently, graph learning models are indispensable tools to help researchers explore graph-structured data. In academia, using sufficient training data to optimize a graph model on a single device is a typical approach for training a capable ...
Event-based fuzzy adaptive control with predetermined performance for MIMO nonlinear systems via nonlinear impulsive dynamics approach
This paper investigates a technical problem, that is, how to apply the backstepping to obtain an event-triggered controller for multiple-input multiple-output (MIMO) nonlinear systems. If the controller is set to be triggered, then all virtual ...
Design and analysis of helper-problem-assisted evolutionary algorithm for constrained multiobjective optimization
In recent years, solving constrained multiobjective optimization problems (CMOPs) by introducing simple helper problems has become a popular concept. To date, no systematic study has investigated the conditions under which this concept operates. ...
Accelerated high-dimensional global optimization: A particle swarm optimizer incorporating homogeneous learning and autophagy mechanisms
The curse of dimensionality often results in either premature convergence or slow convergence in high-dimensional global optimization. This paper proposes the high-speed homogeneous learning-based particle swarm optimizer (HLPSO) to address these ...
Masking and purifying inputs for blocking textual adversarial attacks
The vulnerability of deep neural networks (DNNs) to adversarial attacks has attracted attention in many fields, and researchers have sought methods to improve the robustness of DNNs. Most existing methods are empirical defenses that can only cope ...