Project Details

Hardware-Aware Machine Learning: From Automated Design to Innovative and Explainable Solutions

Project Period: 1. 1. 2024 - 31. 12. 2026

Project Type: grant

Code: GA24-10990S

Agency: Czech Science Foundation

Program: Standardní projekty

Czech title
Strojové učení zohledňující hardware: Od automatizovaného návrhu k inovativním a vysvětlitelným řešením
Type
grant
Keywords

evolutionary algorithm;approximate computing;deep neural network;machine learning;hardware accelerator;explainability;design automation;

Abstract

As machine learning (ML) technology penetrates embedded devices, a new class of design automation algorithms capable of generating hardware-aware implementations of ML algorithms is highly desired. In addition, a lot of effort is now invested in developing explainable ML. We hypothesize that the design time of hardware-aware implementations of ML systems showing additional properties (such as explainable behavior) can be substantially reduced if the used design automation algorithms employ suitable surrogate models for estimating the accuracy, hardware parameters, and other desired properties. In addition to developing suitable surrogate models, we will create a new method based on genetic programming for the automated design of highly-optimized ML models showing excellent trade-offs among the quality of service, hardware parameters, and explainability. The design method and ML models automatically generated by the method will be evaluated in case studies, including image classifiers, Parkinson's disease assessment, and command classifiers of brain signals.

Team members
Sekanina Lukáš, prof. Ing., Ph.D. (UPSY FIT VUT) , research leader
Drahošová Michaela, Ing., Ph.D. (UPSY FIT VUT)
Hurta Martin, Ing. (UPSY FIT VUT)
Malik Aamir Saeed, Ph.D. (UPSY FIT VUT)
Mrázek Vojtěch, Ing., Ph.D. (UPSY FIT VUT)
Piňos Michal, Ing. (UPSY FIT VUT)
Vašíček Zdeněk, doc. Ing., Ph.D. (UPSY FIT VUT)
Zaheer Muhammad Asad (UPSY FIT VUT)
Publications

2024

  • ARIF Muhammad, REHMAN (ur) Faizan, SEKANINA Lukáš and MALIK Aamir Saeed. A comprehensive survey of evolutionary algorithms and metaheuristics in brain EEG-based applications. Journal of Neural Engineering, vol. 21, no. 5, 2024, pp. 1-25. ISSN 1741-2552. Detail
  • PIŇOS Michal, SEKANINA Lukáš and MRÁZEK Vojtěch. ApproxDARTS: Differentiable Neural Architecture Search with Approximate Multipliers. In: 2024 The International Joint Conference on Neural Networks (IJCNN). Yokohama: Institute of Electrical and Electronics Engineers, 2024, pp. 1-8. ISBN 979-8-3503-5931-2. Detail
  • VAŠÍČEK Zdeněk, MRÁZEK Vojtěch and SEKANINA Lukáš. Automated Verifiability-Driven Design of Approximate Circuits: Exploiting Error Analysis. In: 2024 Design, Automation & Test in Europe Conference & Exhibition (DATE). Valencia: Institute of Electrical and Electronics Engineers, 2024, pp. 1-6. ISBN 979-8-3503-4859-0. Detail
  • JAWED Soyiba, FAYE Ibrahima and MALIK Aamir Saeed. Deep learning-based assessment model for Real-time identification of visual learners using Raw EEG. IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 32, no. 1, 2024, pp. 378-390. ISSN 1558-0210. Detail
  • KLHŮFEK Jan, ŠAFÁŘ Miroslav, MRÁZEK Vojtěch, VAŠÍČEK Zdeněk and SEKANINA Lukáš. Exploiting Quantization and Mapping Synergy in Hardware-Aware Deep Neural Network Accelerators. In: 2024 27th International Symposium on Design & Diagnostics of Electronic Circuits & Systems (DDECS). Kielce: Institute of Electrical and Electronics Engineers, 2024, pp. 1-6. ISBN 979-8-3503-5934-3. Detail
  • SEKANINA Lukáš. Tutorial: Evolutionary Design Methods in Electronic Design Automation. In: IEEE 42nd International Conference on Computer Design (ICCD). Milano: IEEE Computer Society, 2024, pp. 689-690. ISBN 979-8-3503-8040-8. Detail
Back to top