Article
Computer Science, Artificial Intelligence
Adhri Nandini Paul, Peizhi Yan, Yimin Yang, Hui Zhang, Shan Du, Q. M. Jonathan Wu
Summary: This research extends two non-iterative training algorithms based on the Moore-Penrose inverse to enable online sequential learning. It uses the proposed autoencoder for self-supervised dimension reduction and the proposed classifier for supervised classification. Experimental results demonstrate that the approach achieves satisfactory classification accuracy on benchmark datasets with extremely low time consumption.
NEURAL COMPUTING & APPLICATIONS
(2021)
Article
Computer Science, Theory & Methods
Piotr S. Maciag, Robert Bembenik, Artur Dubrawski
Summary: In this article, a novel type of spatio-temporal sequential patterns called Constricted Spatio-Temporal Sequential (CSTS) patterns is introduced and analyzed. The CSTS patterns are shown to be a concise representation of all spatio-temporal sequential patterns in a given dataset. A participation index measure is used to determine the significance of the discovered CSTS patterns. An algorithm called CSTS-Miner is provided to discover participation index strong CSTS patterns in event data. Experimental evaluation using crime-related datasets demonstrates that the proposed algorithm discovers much fewer patterns compared to other state-of-the-art algorithms. Examples of interesting crime-related patterns discovered by CSTS-Miner are also presented.
JOURNAL OF BIG DATA
(2023)
Article
Chemistry, Medicinal
Timur Gimadiev, Ramil Nugmanov, Dinar Batyrshin, Timur Madzhidov, Satoshi Maeda, Pavel Sidorov, Alexandre Varnek
Summary: Quantum chemical calculations are widely used for machine learning datasets, but typically lack detailed information on reaction pathways. RePathDB is a database system that manages 3D structural data for both ground and transition states, enabling storage, assembly, and analysis of reaction pathway data. It combines relational and graph database architectures for handling compounds and reactions as molecular graphs.
JOURNAL OF CHEMICAL INFORMATION AND MODELING
(2021)
Article
Computer Science, Artificial Intelligence
Xiumei Wang, Dingning Guo, Peitao Cheng
Summary: Sequential data clustering is a challenging task in data mining, and subspace clustering is a representative tool for dealing with complex local correlation and high-dimensional structure. It is important to learn a more specific structure representation of a sequence to preserve both sequential information and efficient connections.
PATTERN RECOGNITION
(2022)
Review
Automation & Control Systems
Jinjin Zhang, Xiaodong Mu, Peng Zhao, Kai Kang, Chenhui Ma
Summary: The PIRSP model proposed in this paper combines item and review information to learn sequential patterns for sequential recommendation. By selectively combining short-term and long-term preferences using a fusion gating mechanism, the model outperforms other state-of-the-art methods on the Amazon dataset. The analysis of PIRSP's recommendation process shows the positive effect of the two types of information and fusion gating mechanism on the performance of sequential recommendation.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2021)
Article
Computer Science, Information Systems
Zizhuo Zhang, Bang Wang
Summary: This study introduces a new neural model to improve the performance of session-based recommendation task. The model mines items' latent categorical distributions via random walk on an item graph constructed from sessions and consists of two prediction modules: one learns a session's latent categorical representation, the other learns a session's sequential representation. Experimental results demonstrate that our model achieves performance improvements over recent state-of-the-art algorithms on three public datasets.
INFORMATION SCIENCES
(2021)
Article
Computer Science, Information Systems
Chunkai Zhang, Zilin Du, Wensheng Gan, Philip S. Yu
Summary: High-utility sequential pattern mining (HUSPM) has attracted significant research interest recently, with the main task of finding subsequences with high utility in a quantitative sequential database. The top-k HUSPM concept was introduced to address the challenge of specifying a minimum utility threshold. Existing strategies for top-k HUSPM require improvement in terms of efficiency and scalability.
INFORMATION SCIENCES
(2021)
Article
Computer Science, Information Systems
Kashob Kumar Roy, Md Hasibul Haque Moon, Md Mahmudur Rahman, Chowdhury Farhan Ahmed, Carson Kai-Sang Leung
Summary: This paper discusses the increasing importance of imprecise data in the rapid development of science and technology, as well as the challenges in mining patterns in uncertain databases. The authors propose an algorithm to mine frequent sequences in uncertain databases, and introduce two new techniques for incremental mining in databases.
INFORMATION SCIENCES
(2022)
Article
Neurosciences
Jongrok Do, Kang Yong Eo, Oliver James, Joonyeol Lee, Yee-Joon Kim
Summary: It is still unclear how the human brain processes sequential images into abstract representations of mean feature values. In this study, we used multivariate pattern analysis to investigate the sequential averaging mechanism. Our findings suggest that the neural representation of mean orientation becomes increasingly accurate over time and improves more rapidly in stable environments. These findings provide neural mechanisms for accumulating abstract features in different environments.
JOURNAL OF NEUROSCIENCE
(2022)
Article
Chemistry, Physical
Duncan Bossion, Wenxiang Ying, Sutirtha N. Chowdhury, Pengfei Huo
Summary: This paper presents a rigorous theoretical framework for the generalized spin mapping representation and connects previous methods with non-adiabatic quantum dynamics approaches using continuous variables.
JOURNAL OF CHEMICAL PHYSICS
(2022)
Article
Physics, Multidisciplinary
Dominic Else, Ryan Thorngren, T. Senthil
Summary: A system with fractional filling ν imposes strong constraints on the low-energy theory, especially when the system is charge compressible. In such cases, the low-energy theory must have a very large emergent symmetry group. Ersatz Fermi liquids share common properties with Fermi liquids, including periodic quantum oscillations and response to an applied magnetic field.
Article
Automation & Control Systems
Druv Pai, Michael Psenka, Chih-Yuan Chiu, Manxi Wu, Edgar Dobriban, Yi Ma
Summary: In this study, we address the problem of learning discriminative representations for data in high-dimensional space supported on or around multiple low-dimensional linear subspaces. We propose a sequential game approach based on the closed-loop transcription (CTRL) framework, which can learn discriminative and generative representations for general low-dimensional submanifolds. Our theoretical proof demonstrates that the equilibrium solutions to the game provide the correct representations. Our work bridges classical subspace learning methods with modern representation learning techniques and offers the first theoretical justification for the CTRL framework in the case of linear subspaces. We provide empirical evidence and extend the sequential game formulation to more general representation learning problems. Our code is available on GitHub. (c) 2023 The Franklin Institute. Published by Elsevier Inc. All rights reserved.
JOURNAL OF THE FRANKLIN INSTITUTE-ENGINEERING AND APPLIED MATHEMATICS
(2023)
Article
Multidisciplinary Sciences
Dorian Verdel, Simon Bastide, Franck Geffard, Olivier Bruneau, Nicolas Vignais, Bastien Berret
Summary: This study used a robotic exoskeleton to simulate the effects of different gravity fields on the elbow joint. By comparing actual arm movements with predicted results from a control model, it was found that humans can adapt efficiently to a wide range of gravity conditions, opening up new possibilities for rehabilitation and space exploration.
Article
Computer Science, Artificial Intelligence
Dakshi T. Kapugama Geeganage, Yue Xu, Yuefeng Li
Summary: This paper introduces a novel approach to topic representation based on semantic patterns, which considers the semantics and co-occurrence of words to generate frequent semantic patterns for each topic, outperforming in topic quality and information filtering performance evaluations.
KNOWLEDGE-BASED SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Chengxin He, Lei Duan, Guozhu Dong, Jyrki Nummenmaa, Tingting Wang, Tinghai Pang
Summary: Distinguishing sequential patterns are sequences that have higher frequencies in a target group compared to a contrasting group. Previous studies did not consider the hierarchical relationship among elements in sequences. This paper proposes a method to mine distinguishing sequential patterns with concept hierarchies and demonstrates its effectiveness through empirical study.
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Muhammad Aamir Saleem, Rohit Kumar, Toon Calders, Torben Bach Pedersen
KNOWLEDGE AND INFORMATION SYSTEMS
(2019)
Article
Management
Toon Calders, Dimitri Van Assche
OMEGA-INTERNATIONAL JOURNAL OF MANAGEMENT SCIENCE
(2018)
Article
Computer Science, Artificial Intelligence
Karam Gouda, Mona Arafa, Toon Calders
PATTERN RECOGNITION
(2018)
Editorial Material
Computer Science, Artificial Intelligence
Michelangelo Ceci, Toon Calders
Article
Computer Science, Information Systems
Ayman Alserafi, Alberto Abello, Oscar Romero, Toon Calders
ACM TRANSACTIONS ON INFORMATION SYSTEMS
(2020)
Article
Computer Science, Information Systems
Faisal Orakzai, Torben Bach Pedersen, Toon Calders
Summary: The widespread use of mobile devices has generated a large amount of movement data which is being mined to understand collective mobility behaviors of humans, animals, and objects. Convoy pattern is a useful pattern for finding groups moving together. The DCM algorithm proposed in this paper is a scalable and efficient distributed convoy pattern mining algorithm that outperforms existing algorithms.
Article
Computer Science, Interdisciplinary Applications
Hafiz Hassaan Saeed, Muhammad Haseeb Ashraf, Faisal Kamiran, Asim Karim, Toon Calders
Summary: This paper addresses the challenge of Roman Urdu toxic comment detection by developing a first-ever large labeled corpus of toxic and non-toxic comments. With the ensemble approach, the best F1-score reaches 86.35%, setting the first-ever benchmark for toxic comment classification in Roman Urdu.
LANGUAGE RESOURCES AND EVALUATION
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Muhammad Aamir Saleem, Toon Calders, Torben Bach Pedersen, Panagiotis Karras
Summary: This paper explores how to identify and predict groups of companions through social ties and geo-tagged activity information. The proposed nontrivial algorithm COVER uses an activity-driven pruning criterion to guide its exploration and is shown to outperform brute-force baselines and previous work in terms of efficiency and prediction accuracy regarding groups of companions.
WEB AND BIG DATA, APWEB-WAIM 2021, PT II
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Daphne Lenders, Toon Calders
Summary: Situation testing is a method used to prove discrimination by comparing the treatment of similar individuals in the same situation, with the key being finding a suitable distance function to define similarity in the dataset. Recent data-driven equivalents have been proposed, but the challenge lies in determining the appropriate distance function that disregards irrelevant attributes and weighs relevant attributes for classification.
MACHINE LEARNING AND PRINCIPLES AND PRACTICE OF KNOWLEDGE DISCOVERY IN DATABASES, ECML PKDD 2021, PT I
(2021)
Proceedings Paper
Business
Stephen Pauwels, Toon Calders
Summary: Incremental learning strategies are proposed for updating next-activity prediction models for business processes without the need for full retraining, reducing computational resources while maintaining a more consistent and accurate view of the process.
BUSINESS PROCESS MANAGEMENT (BPM 2021)
(2021)
Article
Computer Science, Information Systems
Faisal Orakzai, Toon Calders, Torben Bach Pedersen
PROCEEDINGS OF THE VLDB ENDOWMENT
(2019)
Article
Computer Science, Information Systems
Stephen Pauwels, Toon Calders
APPLIED COMPUTING REVIEW
(2019)
Proceedings Paper
Computer Science, Interdisciplinary Applications
Stephen Pauwels, Toon Calders
SAC '19: PROCEEDINGS OF THE 34TH ACM/SIGAPP SYMPOSIUM ON APPLIED COMPUTING
(2019)
Proceedings Paper
Computer Science, Information Systems
Rohit Kumar, Alberto Abello, Toon Calders
ADVANCES IN DATABASES AND INFORMATION SYSTEMS, ADBIS 2017
(2017)
Proceedings Paper
Computer Science, Artificial Intelligence
Ayman Alserafi, Toon Calders, Alberto Abello, Oscar Romero
SIMILARITY SEARCH AND APPLICATIONS, SISAP 2017
(2017)