Andrzej Cichocki (born 1947) is a Polish computer scientist, electrical engineer and a professor at the Systems Research Institute of Polish Academy of Science, Warsaw, Poland and a visiting professor in several universities and research institutes, especially
Riken AIP, Japan. He is most noted for his learning algorithms for
Signal separation (BSS),
Independent Component Analysis (ICA),
Non-negative matrix factorization (NMF),
tensor decomposition, Deep (Multilayer) Matrix Factorizations for ICA, NMF, PCA, neural networks for optimization and signal processing,
Tensor network for Machine Learning and Big Data, and
brain–computer interfaces. He is the author of several monographs/books [1] and more than 500 scientific peer-reviewed articles.[2]
Education and career
Andrzej Cichocki in 2013 in Riken Brain Science Institute
Andrzej Cichocki received his M.Sc. (with honors), PhD and doctor of science (Dr.Sc.-
Habilitation) degrees all in electrical engineering and computer science from the
Warsaw University of Technology, Poland.
He received the title of full Professor in 1995.
From 1984 to 1989 he was a Alexander von Humboldt Research Fellow and DFG visiting scholar at the University of Erlangen Nurnberg, Germany and he worked closely with Professor
Rolf Unbehauen.
From 1996 till 2018 he worked in RIKEN Brain Science Institute, Wako-shi, Japan at
Shun'ichi Amari's Research Department, as a team leader and later as senior head of laboratories. He established and ran in RIKEN BSI three laboratories: Open Information Systems, Artificial Brains Systems and Cichocki's Laboratory for Advanced Brain Signal Processing.
He pioneered developing and applying new beta and alpha-beta and other divergences in machine learning, especially for non-negative matrix factorizations and nonnegative tensor decompositions. Moreover, he pioneered in development of multilayer (deep) matrix and tensor factorization models and learning algorithms, especially for ICA, NMF and Sparse Component Analysis (SCA).[6][7][8]
He developed and proposed new recurrent neural network architectures for optimization, solving large scale systems of algebraic equations and blind signal separation, especially multilayer (deep) hierarchical neural networks. He contributed to development of natural gradient algorithms for Independent Component Analysis (ICA) and blind deconvolution.[9][10]
He proposed together with his co-workers several efficient AI models and machine learning algorithms for brain computer interface, human emotions recognition and early diagnosis of some brain diseases, like Alzheimer and Schizophrenia.
Following concerns raised by some AI experts about the potential risks that AGI may pose on humanity, Cichocki suggested in 2021 development of novel AGI systems with implemented multiple intelligences, including not only ethical/moral intelligence but also social-emotional intelligence with self-awareness and responsible decision making abilities.
His current research interests include:
Tensor decomposition and tensor networks
Learning of non-stationarity data
Data fusion of multi-modal structured data, and deep neural networks compression
Cichocki, A., & Amari, S. I. (2002). Adaptive Blind Signal and Image Processing: Learning Algorithms and Applications. John Wiley & Sons.
ISBN9780470845899,
doi:
10.1002/0470845899
Cichocki, A., Zdunek, R., Phan, A. H., & Amari, S. I. (2009). Nonnegative Matrix and Tensor Factorizations: Applications to Exploratory Multi-way Data Analysis and Blind Source Separation. John Wiley & Sons.
ISBN9780470747278,
doi:
10.1002/9780470747278
Cichocki, A., Lee, N., Oseledets, I., Phan, A. H., Zhao, Q., & Mandic, D. P. (2016). Tensor Networks for Dimensionality Reduction and Large-Scale Optimization: Part 1 Low-Rank Tensor Decompositions. Foundations and Trends® in Machine Learning, 9(4-5), 249-429.
doi:
10.1561/2200000059
Cichocki, A., Phan, A. H., Zhao, Q., Lee, N., Oseledets, I., Sugiyama, M., & Mandic, D. P. (2017). Tensor Networks for Dimensionality Reduction and Large-Scale Optimization: Part 2 Applications and Future Perspectives. Foundations and Trends® in Machine Learning, 9(6), 431-673.
doi:
10.1561/2200000067
2018 The best paper award in 2018 in IEEE Signal Processing Magazine for the paper “Tensor decompositions for signal processing applications: From two-way to multiway component analysis”, coauthored by A. Cichocki, D. Mandic, L De Lathauwer, A.H. Phan, Q. Zhao, C. Caiafa, G, Zhao [13]
2018 H.C. (Honoris Causa) Doctorate, awarded by Nicolaus Copernicus University, Torun, Poland, February 27, 2022 [14]
2016 Excellent ICONIP Paper Award for the paper authored by Namgil Lee, Anh-Huy Phan, Fengyu Cong, Andrzej Cichocki. “Nonnegative tensor train decompositions for multi-domain feature extraction and clustering”
2015 The best paper award in Journal Entropy the paper “Generalized Alpha-Beta divergences and their application to robust non-negative matrix factorization” Entropy 2011, 13(1), 134–170; coauthored by A. Cichocki, S. Cruces and S. Amari [15]
2014 The Best paper award in Journal Entropy for 2014 for the paper coauthored by Andrzej Cichocki and
Shun'ichi Amari, “Families of Alpha- Beta- and Gamma- Divergences: Flexible and robust measures of similarities” [16]
2010 APNNA Best Paper Award for the paper coauthored by Yunjun Nam, Qibin Zhao, Andrzej Cichocki, and Seungjin Choi “A tongue-machine interface: Detection of tongue positions by glossokinetic potentials,” in Proceedings of the International Conference on Neural Information Processing (ICONIP-2010), Sydney, Australia, November 22–25, 2010.
1995 Received a title of Professor in Poland from the President of the country
^Amari, Shun'ichi; Cichocki, Andrzej; Young, Howard (1995).
"A new learning algorithm for blind signal separation"(PDF). Proc. Advances in Neural Information Processing Systems 8 Advances in Neural Information Processing Systems 8: 757–763.