TETRA-AI | NMBU

Ordsky med fokus på "Explainable AI", der de mest fremtredende ordene inkluderer 'Explainable', 'AI', 'Human Understanding' og 'Global Explanation'. Andre viktige begreper er 'Transparency', 'Interpretability', 'Ethical AI' og 'Local Explanation'. Bakgrunnen har en blå fargegradient fra mørk øverst til lys nederst." — **Photo:** Shutterstock

Trustworthy and Efficient Transfer Learning AI (TETRA AI) is a research group dedicated to advancing methods that make AI models both trustworthy and highly efficient.

Faculty

Faculty of Science and Technology

Shortcuts

About the research group

Research areas
The group focuses on developing novel techniques such as low-rank approximation to reduce model complexity and computational cost, while injecting formal domain knowledge to enhance transparency, robustness, and trustworthiness. TETRA AI’s research enables zero-shot and few-shot learning, empowering models to adapt reliably to new tasks with minimal labeled data. The team’s mission is to create trustworthy, resource-efficient AI systems that can be confidently deployed in real-world, high-stakes environments.

Members

Scientific staff
PhD candidates and post doctors
- External partners and PhD candidates
  Univ.-Prof. Dr.-Ing. Kyandoghere Kyamakya
  https://www.aau.at/en/team/kyamakya-kyandoghere/
  Univ.-Prof. Dr. Martin Gebser
  https://www.aau.at/en/aics/research-groups/research-group-production-systems/team/martin-gebser/
  Ali Deeb (stipendiat)
  Vianney Kambale (stipendiat)

Projects

Publications: Efficient Transfer Learning

FragFM: Fragmented Fourier Matrix for multi-seasonality extraction in Univariate Long-Term Time Series Forecasting

Aditya Dey, Jonas Kusch, Fadi Al Machot

Norwegian ICT Conference for Research and Education, 2025.

“SEER-ZSL: Semantic Encoder-Enhanced Representations for Generalized Zero-Shot Learning”

William Heyden, Habib Ullah, M.S. Siddiqui, Fadi Al-Machot

Proceedings of the Winter Conference on Applications of Computer Vision (WACV), pp. 30–41, 2025.

“RevCD – Reversed Conditional Diffusion for Generalized Zero-Shot Learning”

William Heyden, Habib Ullah, M.S. Siddiqui, Fadi Al-Machot

International Conference on Deep Learning Theory and Applications (DeLTA), 2024.

“HADL: A Haar-DCT and Low-Rank Framework for Noise-Resilient Long-Term Time Series Forecasting”

Aditya Dey, Jonas Kusch, Fadi Al Machot

International Conference on Time Series and Forecasting (ITISE), 2025.

“An Integral Projection-Based Semantic Autoencoder for Zero-Shot Learning”

William Heyden, Habib Ullah, M.S. Siddiqui, Fadi Al-Machot

IEEE Access 2023

CLAW: Complex Low-rank and Wavelet Transform for Univariate Long-term Time Series Forecasting

Aditya Dey, Jonas Kusch, and Fadi Al Machot

International Conference on Frontiers of Artificial Intelligence, Ethics, and Multidisciplinary Applications (FAIEMA ), 2025.

Lightweight and Resource-Efficient Battery Degradation Prediction Using Teacher-Informed Cellular Neural Networks (TICeNN)

Liyanapathiranage Sudeepika Wajirakumari Samarathunga, Aditya Dey, Martin Thomas Horsch, and Fadi Al Machot

International Conference on Frontiers of Artificial Intelligence, Ethics, and Multidisciplinary Applications (FAIEMA ), 2025.

Enhanced Time Series Forecasting: Integrating PatchTST with BERT Layers

Hussein Ahmad Ahmad, Seyyed Kasra Mortazavi, Mohamed El Bahnasawi, Fadi Al Machot, Witesyavwirwa Vianney Kambale, Kyandoghere Kyamakya

IEEE International Conference on Applied Mathematics & Computer Science (ICAMCS), 2024.

Publications: Trustworthy AI, Explainable AI, and Knowledge-Infused Learning

“Symbolic-AI-Fusion Deep Learning (SAIF-DL): Encoding Knowledge into Training with Answer Set Programming Loss Penalties by a Novel Loss Function Approach”

Fadi Al Machot, Martin Thomas Horsch, Habib Ullah

Designing the Conceptual Landscape for a XAIR Validation Infrastructure, Springer Nature, 2025.

“Building Trustworthy AI: Transparent AI Systems via Language Models, Ontologies, and Logical Reasoning (TranspNet)”

Fadi Al Machot, Martin Thomas Horsch, Habib Ullah

Designing the Conceptual Landscape for a XAIR Validation Infrastructure, Springer Nature, 2025.

“Bridging Logic and Learning: A Neural-Symbolic Approach for Enhanced Reasoning in Neural Models (ASPER)”

Fadi Al Machot

Designing the Conceptual Landscape for a XAIR Validation Infrastructure, Springer Nature, 2025.

“ NASP-T: A Fuzzy Neuro-Symbolic Transformer for Logic-Constrained Aviation Safety Report Classification ”

Fadi Al Machot, Fidaa Al Machot

arXiv, 2025.

“Towards Explainable Ensemble Learning: Adaptive Feature Space Pruning and Optimized Classifier Selection”

Muhammad Uzair, Zohaib Jan, Sultan Daud Khan, Fadi Al Machot, Habib Ullah

Designing the Conceptual Landscape for a XAIR Validation Infrastructure, Springer Nature, 2025.

Publications: Robust and Generalizable Neural Architectures for Knowledge Extraction

Genomic prediction of feed efficiency in boars by deep learning

Olumide Onabanjo, Theo Meuwissen, Hans Magnus Gjøen, Fadi Al Machot, Maren van Son, Peer Be

G3: Genes, Genomes, Genetics, 2026

Towards sustainable universal deepfake detection with frequency-domain masking

Chandler Timm C Doloriel, Habib Ullah, Kristian Hovde Liland, Fadi Al Machot, Ngai-Man Cheung

ACM Transactions on Multimedia Computing, Communications and Applications, 2025

“A Robust Automated Analog Circuits Classification Involving a Graph Neural Network and a Novel Data Augmentation Strategy”

Ali Deeb, Abdalrahman Ibrahim, Mohamed Salem, Joachim Pichler, Sergii Tkachov, Anjeza Karaj, Fadi Al Machot, Kyamakya Kyandoghere

Sensors, 2023.

“A Graph Attention Network Based System for Robust Analog Circuits’ Structure Recognition Involving a Novel Data Augmentation Technique”

Ali Deeb, Mohamed Salem, Abdalrahman Ibrahim, Joachim Pichler, Sergii Tkachov, Anjeza Karaj, Fadi Al Machot, Kyandoghere Kyamakya

IEEE Access, 2024.

Relevant study programs

Master in Data Science (5 years)

Master in Industrial Economy (5 years)

Master in Applied Robotics (5 years)

About the research group

Research areas

Members

Scientific staff

PhD candidates and post doctors

External partners and PhD candidates

Projects

Publications: Efficient Transfer Learning

FragFM: Fragmented Fourier Matrix for multi-seasonality extraction in Univariate Long-Term Time Series Forecasting

“SEER-ZSL: Semantic Encoder-Enhanced Representations for Generalized Zero-Shot Learning”

“RevCD – Reversed Conditional Diffusion for Generalized Zero-Shot Learning”

“HADL: A Haar-DCT and Low-Rank Framework for Noise-Resilient Long-Term Time Series Forecasting”

“An Integral Projection-Based Semantic Autoencoder for Zero-Shot Learning”

CLAW: Complex Low-rank and Wavelet Transform for Univariate Long-term Time Series Forecasting

Lightweight and Resource-Efficient Battery Degradation Prediction Using Teacher-Informed Cellular Neural Networks (TICeNN)

Enhanced Time Series Forecasting: Integrating PatchTST with BERT Layers

Publications: Trustworthy AI, Explainable AI, and Knowledge-Infused Learning

“Symbolic-AI-Fusion Deep Learning (SAIF-DL): Encoding Knowledge into Training with Answer Set Programming Loss Penalties by a Novel Loss Function Approach”

“Building Trustworthy AI: Transparent AI Systems via Language Models, Ontologies, and Logical Reasoning (TranspNet)”

“Bridging Logic and Learning: A Neural-Symbolic Approach for Enhanced Reasoning in Neural Models (ASPER)”

“ NASP-T: A Fuzzy Neuro-Symbolic Transformer for Logic-Constrained Aviation Safety Report Classification ”

“Towards Explainable Ensemble Learning: Adaptive Feature Space Pruning and Optimized Classifier Selection”

Publications: Robust and Generalizable Neural Architectures for Knowledge Extraction

Genomic prediction of feed efficiency in boars by deep learning

Towards sustainable universal deepfake detection with frequency-domain masking

“A Robust Automated Analog Circuits Classification Involving a Graph Neural Network and a Novel Data Augmentation Strategy”

“A Graph Attention Network Based System for Robust Analog Circuits’ Structure Recognition Involving a Novel Data Augmentation Technique”

Relevant study programs