## KYUSHU UNIVERSITY |
## Graduate School and Faculty of Information Science and Electrical Engineering |
||

## Deparment of Electrical Engineering and Computer Science, School of Engineering |

## About us

Mathematical Engineering Laboratory offers education and research on various circuits for information transmission, analog/digital signal processing, and modern digital communication systems, based on circuit theory, nonlinear theory, information theory, probability theory and communication theory. Main research topics of this laboratory are as follows: 1) analysis and design methods associated with large scale integrated circuits, 2) theoretical and experimental clarification of complicated phenomena in nonlinear circuits including neural networks as well as their application to signal processing technology, 3) construction of digital chaos systems based on nonlinear theory and their application to digital communication networks, 4) basic problems in information theory and related areas.

## News & Topics

**2018.3****2018.3****(3/19) The following paper has been published in Computational Optimization and Applications**- Norikazu Takahashi (Okayama Univ.), Jiro Katayama (OB of our Lab), Masato Seki (Okayama Univ.)，Jun’ichi Takeuchi (Kyushu Univ.), "A unified global convergence analysis of multiplicative update rules for nonnegative matrix factorization"

**(3/5) Professor Jun'ichi Takeuchi had a presentation at Technical Committee on Infomation-Based Induction Sciences and Machine Learning (IBISML)**- Junichi Takeuchi, Hiroshi Nagaoka (University of Electro-Communications), "Exponential Family of Markov Kernels and Asymptotic Exponential Family of Markov Sources"

**2017.11****(11/9) Professor Jun'ichi Takeuchi had a presentation at****IEEE Information Theory Workshop(ITW2017)**- Junichi Takeuchi, Hiroshi Nagaoka (University of Electro-Communications), "Information Geometry of The Family of Markov Kernels Defined by A Context Tree"

**2017.9****(9/8) Munenori Eto(M1) had a presentation at Technical Committee Conference on Information Theory, IEICE**- Munenori Eto, Masanori Kawakita, Jun'ichi Takeuchi， "Cardinality of Typical Set and the Smallest High Probability Set for Discrete Memoryless Sources and Markov Sources"

**2017.7****(7/21) The following paper has been accepted for IEEE Information Theory Workshop(ITW2017)**- Jun'ichi Takeuchi, Hiroshi Nagaoka（UEC）, "Information Geometry of The Family of Markov Kernels Defined by A Context Tree"

**(7/14) Professor Jun'ichi Takeuchi had presentations at Technical Committee Conference on Information Theory, IEICE**- Jun'ichi Takeuchi, Hiroshi Nagaoka（UEC）， "On Information Geometry of Tree Models "

- Jun'ichi Takeuchi, Hiroshi Nagaoka（UEC）， "Asymptotic Exponential Family of Markov Sources is Equivalent to Exponential Family of Markov Kernelss "

**2017.6****(6/28) Associate Professor Yutaka Jitsumatsu had a presentation at IEEE International Symposium on Information Theory (ISIT2017)**- Yutaka Jitsumatsu，"Computation of the Random Coding Secrecy Exponent for a Constant Composition Ensemble"

**(6/28) Dongshin Yang (D3 student) had a presentation at International Radar Symposium 2017 (IRS 2017)**- Dongshin Yang, Yutaka Jitsumatsu，"Compressive Sensing of Up-sampled Model and Atomic Norm for Super-Resolution Rader"

**(6/22) The following paper has been pubulished on Machine Learning**- Masanori Kawakita, Jun'ichi Takeuchi, "A note on model selection for small sample regression"

**2017.4****(4/2) The following paper has been accepted for IEEE International Symposium on Information Theory (ISIT2017)**- Yutaka Jitsumatsu，"Computation of the Random Coding Secrecy Exponent for a Constant Composition Ensemble"

**2017.2****(2/19) The following paper has been accepted for International Radar Symposium 2017 (IRS 2017)**- Dongshin Yang, Yutaka Jitsumatsu，"Compressive Sensing of Up-sampled Model and Atomic Norm for Super-Resolution Rader"

**2016.10****(10/18) Shoma Tanaka(M2) and Chansu Han(M1) had a presentation at 2016 Data Mining and Cybersecurity Workshop, associated with ICONIP 2016**- Shoma Tanaka, Yuki Kawamura, Masanori Kawakita, Noboru Murata, and Jun'ichi Takeuchi, "MDL Criterion for NMF with Application to Botnet Detection"
- Chansu Han, Kento Kono, Shoma Tanaka, Masanori Kawakita, and Jun'ichi Takeuchi, "Botnet Detection Using Graphical Lasso with Graph Density"

**2016.8****(8/26) Dongshin Yang, D2 student, had a presentation at The 13th IEEE VTS Asia Pacific Wireless Communications Symposium (APWCS 2016)**- Dongshin Yang, Yutaka Jitsumatsu， "Discretization in Channel Estimation Using Compressed Sensing and Its Performance Improvement"

**2016.7****(7/12,7/13) Yoshinari Takeishi and Associate Professor Yutaka Jitsumatsu had a presentation at 2016 IEEE International Symposium on Information Theory (ISIT2016)**- (7/13)Yutaka Jitsumatsu, Yasutada Oohama (University of Electro- Communications), "Computing the Optimal Exponent of Correct Decoding for Discrete Memoryless Sources"
- (7/12)Yoshinari Takeishi, Jun'ichi Takeuchi, "An Improved Upper Bound on Block Error Probability of Least Squares Superposition Codes with Unbiased Bernoulli Dictionary"

**(7/7) The following papers have been accepted for 2016 Data Mining and Cybersecurity Workshop, associated with ICONIP 2016**- Shoma Tanaka, Yuki Kawamura, Masanori Kawakita, Noboru Murata, and Jun'ichi Takeuchi, "MDL criterion for NMF with Application to Botnet Detection"
- Chansu Han, Kento Kono, Shoma Tanaka, Masanori Kawakita, and Jun'ichi Takeuchi, "Botnet Detection Using Graphical Lasso with Graph Density"

**(7/1) The following papers have been accepted for 2016 International Symposium on Nonlinear Theory and Its Applications(NOLTA2016)**- Yeelai Chew, Yutaka Jitsumatsu, "Estimation of Beta-Value for Pipelined Beta Encoders"
- Koji Itaya, Yutaka Jitsumatsu, "Random Number Generation Using Outputs from Multiple Beta Encoders"

**2016.6****(6/20) Assistant Professor Masanori Kawakita had a presentation at The 33rd International Conference on Machine Learning (ICML 2016)**- Masanori Kawakita, Jun'ichi Takeuchi, "Barron and Cover's Theory in Supervised Learning and its Application to Lasso"

**(6/15) The following paper has been accepted for The 13th IEEE VTS Asia Pacific Wireless Communications Symposium (APWCS 2016)**- Dongshin Yang, Yutaka Jitsumatsu， "Discretization in Channel Estimation Using Compressed Sensing and Its Performance Improvement"

**2016.4****(4/24) The following paper has been accepted for The 33rd International Conference on Machine Learning (ICML 2016)**- Masanori Kawakita, Jun'ichi Takeuchi, "Barron and Covers' Theory in Supervised Learning and Its Application to Lasso"

**(4/3) The following papers have been accepted for 2016 IEEE International Symposium on Information Theory (ISIT2016)**- Yutaka Jitsumatsu, Yasutada Oohama (University of Electro- Communications), "Computing the Optimal Exponent of Correct Decoding for Discrete Memoryless Sources"
- Yoshinari Takeishi, Jun'ichi Takeuchi, "An Improved Upper Bound on Block Error Probability of Least Squares Superposition Codes with Unbiased Bernoulli Dictionary"

∗ Yoshinari Takeishi was a master course student of our laboratory

**2016.1****(1/18) Yoshinari Takeishi had a presentation at Technical Committee Conference on Information Theory, IEICE**- Yoshinari Takeishi, Jun'ichi Takeuchi, "An Improved Upper Bound on Block Error Probability of Least Sauares Sparse Superposition Codes with Bernoulli Dictionary"

**(1/1) The following paper was pubulished on Nonlinear Theory and Its Applications, IEICE**- Yutaka Jitsumatsu, Kazuya Matsumura, "A β-ary to binary conversion for random number generation using a β encoder"

∗ Kazuya Matsumura was a master course student of our laboratory

- Yutaka Jitsumatsu, Kazuya Matsumura, "A β-ary to binary conversion for random number generation using a β encoder"
**2015.12****(12/3) Yeelai Chew had a presentation at 2015 International Symposium on Nonlinear Theory and its Applications (NOLTA2015)and received Student Paper Award**- Yeelai Chew, Yutaka Jitsumatsu, Kohei Oda, "Performance Evaluation of a Random Number Generation Using a Beta Encoder"

**2015.11****(11/27) Associate Professor Yutaka Jitsumatsu had a presentation at SITA 2015**- Yutaka Jitsumatsu, Yasutada Oohama (University of Electro-Communications), "A New Iterative Algorithm for Computing the Optimal Exponent of Correct Decoding Probability for Discrete Memoryless Channels under Input Constraint"

**2015.10****Dongshin Yang had a presentation at Math for Industry Forum 2015**- Dongshin Yang, Yutaka Jitsumatsu, "Channel Estimation and Super Resolution Radar by Using Compressed Sensing"

**2015.9****Shutaro Ito had a presentation at 2015 IEEE 82nd Vehicular Technology Conference: VTC2015-Fall**- Shuntaro Ito, Dongshin Yang, Yutaka Jitsumatsu, "Estimation of Multi-Path Channnels by Using the Annihilating Filter Method"

**2015.9****Associate Professor Yutaka Jitsumatsu had a presentation at 2015 IEEE International Symposium on Information Theory (ISIT2015)**- Yasutada Oohama (University of Electro-Communications), Yutaka Jitsumatsu, "A New Iterative Algorithm for Computing the Optimal Exponent of Correct Decodeing for Discrete Memoryless Channel"