Kenji Sugimoto

Chengyan Zhao, Masaki Ogura, Kenji Sugimoto

In this paper, we study the problem of optimizing the stability of positive semi-Markov jump linear systems. We specifically consider the problem of tuning the coefficients of the system matrices for maximizing the exponential decay rate of the system under a budget-constraint. By using a result from the matrix theory on the log-log convexity of the spectral radius of nonnegative matrices, we show that the stability optimization problem reduces to a convex optimization problem under certain regularity conditions on the system matrices and the cost function. We illustrate the validity and effectiveness of the proposed results by using an example from the population biology.

Wendyam Eric Lionel Ilboudo, Taisuke Kobayashi, Kenji Sugimoto

Machine learning algorithms aim to find patterns from observations, which may include some noise, especially in robotics domain. To perform well even with such noise, we expect them to be able to detect outliers and discard them when needed. We therefore propose a new stochastic gradient optimization method, whose robustness is directly built in the algorithm, using the robust student-t distribution as its core idea. Adam, the popular optimization method, is modified with our method and the resultant optimizer, so-called TAdam, is shown to effectively outperform Adam in terms of robustness against noise on diverse task, ranging from regression and classification to reinforcement learning problems. The implementation of our algorithm can be found at https://github.com/Mahoumaru/TAdam.git

Wendyam Eric Lionel Ilboudo, Taisuke Kobayashi, Kenji Sugimoto

Behavioral cloning (BC) bears a high potential for safe and direct transfer of human skills to robots. However, demonstrations performed by human operators often contain noise or imperfect behaviors that can affect the efficiency of the imitator if left unchecked. In order to allow the imitators to effectively learn from imperfect demonstrations, we propose to employ the robust t-momentum optimization algorithm. This algorithm builds on the Student's t-distribution in order to deal with heavy-tailed data and reduce the effect of outlying observations. We extend the t-momentum algorithm to allow for an adaptive and automatic robustness and show empirically how the algorithm can be used to produce robust BC imitators against datasets with unknown heaviness. Indeed, the imitators trained with the t-momentum-based Adam optimizers displayed robustness to imperfect demonstrations on two different manipulation tasks with different robots and revealed the capability to take advantage of the additional data while reducing the adverse effect of non-optimal behaviors.

Toshihide Tadenuma, Masaki Ogura, Kenji Sugimoto

In this paper, we study the problem of continuous-time state observation over lossy communication networks. We consider the situation in which the samplers for measuring the output of the plant are spatially distributed and their communication with the observer is scheduled according to a round-robin scheduling protocol. We allow the observer gains to dynamically change in synchronization with the scheduling of communications. In this context, we propose a linear matrix inequality (LMI) framework to design the observer gains that ensure the asymptotic stability of the error dynamics in continuous time. We illustrate the effectiveness of the proposed methods by several numerical simulations.