Outputs (20)

A Comparative Study of Assessment Metrics for Imbalanced Learning (2023)
Presentation / Conference Contribution
Farou, Z., Aharrat, M., & Horváth, T. (2023, September). A Comparative Study of Assessment Metrics for Imbalanced Learning. Presented at European Conference on Advances in Databases and Information Systems (ADBIS 2023), Barcelona, Spain

There are several machine learning algorithms addressing class imbalance problem, requiring standardized metrics for adequete performance evaluation. This paper reviews several metrics for imbalanced learning in binary and multi-class problems. We em... Read More about A Comparative Study of Assessment Metrics for Imbalanced Learning.

Squared Symmetric Formal Contexts and Their Connections with Correlation Matrices (2023)
Presentation / Conference Contribution
Antoni, L., Eliaš, P., Horváth, T., Krajči, S., Krídlo, O., & Török, C. (2023, September). Squared Symmetric Formal Contexts and Their Connections with Correlation Matrices. Presented at International Conference on Conceptual Structures (ICCS) 202

Formal Concept Analysis identifies hidden patterns in data that can be presented to the user or the data analyst. We propose a method for analyzing the correlation matrices based on Formal concept analysis. In particular, we define a notion of square... Read More about Squared Symmetric Formal Contexts and Their Connections with Correlation Matrices.

NCC: Neural concept compression for multilingual document recommendation (2023)
Presentation / Conference Contribution
Tashu, T. M., Lenz, M., & Horváth, T. NCC: Neural concept compression for multilingual document recommendation

In this work, we propose a novel method for generating inter-lingual document representations using neural network concept compression. The presented approach is intended to improve the quality of content-based multilingual document recommendation an... Read More about NCC: Neural concept compression for multilingual document recommendation.

Hyper-parameter initialization of classification algorithms using dynamic time warping: A perspective on PCA meta-features (2022)
Presentation / Conference Contribution
Horváth, T., Mantovani, R. G., & de Carvalho, A. C. Hyper-parameter initialization of classification algorithms using dynamic time warping: A perspective on PCA meta-features

Meta-learning, a concept from the area of automated machine learning, aims at providing decision support for data scientists by recommending a suitable setting (a machine learning algorithm or its hyper-parameters) to be used for a given dataset. Suc... Read More about Hyper-parameter initialization of classification algorithms using dynamic time warping: A perspective on PCA meta-features.

Solving Multi-class Imbalance Problems Using Improved Tabular GANs (2022)
Presentation / Conference Contribution
Farou, Z., Kopeikina, L., & Horváth, T. (2022, November). Solving Multi-class Imbalance Problems Using Improved Tabular GANs. Presented at 23rd International Conference on Intelligent Data Engineering and Automated Learning (IDEAL), Manchester

Multi-class imbalance problems are non-standard derivative data science problems. These problems are associated with the skewness in the data underlying distribution, which, in turn, raises numerous issues for conventional machine learning techniques... Read More about Solving Multi-class Imbalance Problems Using Improved Tabular GANs.

Synonym-Based Essay Generation and Augmentation for Robust Automatic Essay Scoring (2022)
Presentation / Conference Contribution
Tashu, T. M., & Horváth, T. (2022, November). Synonym-Based Essay Generation and Augmentation for Robust Automatic Essay Scoring. Presented at 23rd International Conference on Intelligent Data Engineering and Automated Learning (IDEAL), Manchester

Automatic essay scoring (AES) models based on neural networks (NN) have had a lot of success. However, research has shown that NN-based AES models have robustness issues, such that the output of a model changes easily with small changes in the input.... Read More about Synonym-Based Essay Generation and Augmentation for Robust Automatic Essay Scoring.

Denoising Architecture for Unsupervised Anomaly Detection in Time-Series (2022)
Presentation / Conference Contribution
Skaf, W., & Horváth, T. (2022, September). Denoising Architecture for Unsupervised Anomaly Detection in Time-Series. Presented at ADBIS 2022: 26th European Conference on Advances in Databases and Information Systems, Turin, Italy

Anomalies in time-series provide insights of critical scenarios across a range of industries, from banking and aerospace to information technology, security, and medicine. However, identifying anomalies in time-series data is particularly challenging... Read More about Denoising Architecture for Unsupervised Anomaly Detection in Time-Series.

Directed Undersampling Using Active Learning for Particle Identification (2022)
Presentation / Conference Contribution
Farou, Z., Ouaari, S., Domian, B., & Horváth, T. (2021, May). Directed Undersampling Using Active Learning for Particle Identification. Presented at 4th International Conference on Recent Innovations in Computing (ICRIC-2021), Central University of Jam

Migrating Models: A Decentralized View on Federated Learning (2021)
Presentation / Conference Contribution
Kiss, P., & Horváth, T. (2021, September). Migrating Models: A Decentralized View on Federated Learning. Presented at ECML PKDD 2021, Online

Federated learning (FL) researches attempt to alleviate the increasing difficulty of training machine learning models, when the training data is generated in a massively distributed way. The key idea behind these methods is moving the training to loc... Read More about Migrating Models: A Decentralized View on Federated Learning.

Time-Series in Hyper-parameter Initialization of Machine Learning Techniques (2021)
Presentation / Conference Contribution
Horváth, T., Mantovani, R. G., & de Carvalho, A. C. P. L. F. (2021, November). Time-Series in Hyper-parameter Initialization of Machine Learning Techniques. Presented at 22nd International Conference on Intelligent Data Engineering and Automated Learning

Initializing the hyper-parameters (HPs) of machine learning (ML) techniques became an important step in the area of automated ML (AutoML). The main premise in HP initialization is that a HP setting that performs well for a certain dataset(s) will als... Read More about Time-Series in Hyper-parameter Initialization of Machine Learning Techniques.