Genomic selection (GS) is expected to accelerate plant and animal breeding. During the last decade, genome-wide polymorphism data have increased, which has raised concerns about storage cost and computational time. Several individual studies have attempted to compress the genome data and predict phenotypes. However, compression models lack adequate quality of data after compression, and prediction models are time consuming and use original data to predict the phenotype. Therefore, a combined application of compression and genomic prediction modeling using deep learning could resolve these limitations. A Deep Learning Compression-based Genomic Prediction (DeepCGP) model that can compress genome-wide polymorphism data and predict phenotypes of a target trait from compressed information was proposed. The DeepCGP model contained two parts: (i) an autoencoder model based on deep neural networks to compress genome-wide polymorphism data, and (ii) regression models based on random forests (RF), genomic best linear unbiased prediction (GBLUP), and Bayesian variable selection (BayesB) to predict phenotypes from compressed information. Two datasets with genome-wide marker genotypes and target trait phenotypes in rice were applied. The DeepCGP model obtained up to 99% prediction accuracy to the maximum for a trait after 98% compression. BayesB required extensive computational time among the three methods, and showed the highest accuracy; however, BayesB could only be used with compressed data. Overall, DeepCGP outperformed state-of-the-art methods in terms of both compression and prediction. Our code and data are available at https://github.com/tanzilamohita/DeepCGP.

This paper addresses the modeling and measurement of a small boat. In some fishing tasks, anchorage is not applicable in order to capture shellfishes or fishes efficiently. Currently, many fishermen are manually stabilizing boats simultaneously with the capturing task. We propose a boat modeling method named FAD learning. In this method, the free dynamics acceleration and actuator acceleration of a boat are learned using the online learning of two dynamics learning trees (DLTs), which are developed by us. In order to measure the position, velocity, and acceleration, we developed an image processing method with an underwater camera. In the experiment, the motor babbling of a boat was performed on a water pool. The dynamical data from the boat was learned by DLTs. The effectiveness of the modeling was confirmed through the validation of the velocity that was predicted by DLTs.

A variety of real-world problems are characterized by multimodal functions, i.e., they demand the identification of several locally or globally optimal solutions. In the field of multimodal optimization, researchers actively address the development of niching techniques by inventing new methods and also by coupling existing methods with global optimization metaheuristics. In this paper, we put forward a general niching framework for the Variable Mesh Optimization (VMO) metaheuristic algorithm, coined as Variable Mesh Optimization with Niching (VMO-N). We argue that the basic VMO formulation can be easily adapted to multimodal optimization scenarios without extra adjustments. We refer to the recently published Niche-Clearing-based Variable Mesh Optimization (NC-VMO), as an instantiation of the VMO-N framework proposed here. Since the niche clearing procedure in NC-VMO requires a priori knowledge of the fitness landscape, we also introduce the Niche-based VMO via Adaptive Species Discovery (VMO-ASD) as another instance of VMO-N, in order to show its applicability to a wider range of multimodal problems. VMOASD results with some of the benchmark functions for the CEC'2013 Special Session and Competition on Niching Methods for Multimodal Function Optimization confirm that VMO-N is a plausible general VMO niching framework for multimodal problems. We demonstrate the advantages of VMO-N over the canonical version of the algorithm, VMO-ASD over NC-VMO and also report on VMO-N's performance against other state-of-the-art niching algorithms. Further pursuits of strongly competitive versions of VMO-N will demand the use of more suitable niching methods than both clearing and ASD within the framework although this is not the focal point of this study.

Currently, many studies have been conducted on robot interactions with humans. Object recognition and feature extraction are essential functions for such robots. Discernment behavior is a type of exploratory behavior that supports object feature extraction. We have proposed an active perception model that autonomously learns discernment behaviors. We have shown the effectiveness of our model using a mobile robot simulation. In this study, we applied our model to a real humanoid robot and confirmed that the robot successfully learns exploratory behaviors. We show that the robot can learn suitable exploratory behaviors by online learning applicable to real-world environments.

This paper addresses the rhythmic reference in physical human-robot interaction. Human refers to a rhythm from multiple sensing modalities when turning a rope with another human synchronously. This study verifies a hypothesis that some humans mix several rhythms of the modalities into a rhythm (rhythmic reference). Six participants, four males and two females, 21-23 years old, took part in eight experiments which examined the hypothesis. In each experiment, we masked the perception of each participant using eight combination of three kinds of masks, an eye-mask, headphones, and a force mask. Each participant interacted with an operator that turned a rope with a constant frequency. As a result of the experiments, a participant increased the controlling error as the number of masks was increased regardless the types of masked modalities. The result strongly supported our hypothesis.

This paper presents a general method for creating an approximately optimal online stabilization system. An optimal stabilization system is an ideal online system that can calculate each optimal motion leading to a stable mechanical goal state depending on the current state. We propose a system that selects each semi-optimal motion according to the current state from a reverse-time tree. To create the reverse-time tree, we applied rapid semi-optimal motion planning method (RASMO) to a reverse-time search from a stable state. We also developed an online motion selection technique. To validate the proposed method, we simulated the stabilization of a double inverted pendulum. When we used an optimization criteria, time optimal, the system quickly stabilized the pendulum's posture and velocity. When we used higher resolution RASMO, the time approached the optimal time. The general framework proposed here is applicable to a variety of machines.

We have proposed self-organizing network elements (SONE) as a learning method for robots to meet the requirements of autonomous exploration of effective output, simple external parameters, and low calculation costs. SONE can be used as an algorithm for obtaining network topology by propagating reinforcement signals between the elements of a network. Traditionally, the analysis of fundamental features in SONE and their application to supervised learning tasks were difficult because the learning method of SONE was limited to reinforcement learning. Here the abilities of generalization, incremental learning, and temporal sequence learning were evaluated using a supervised learning method with SONE. Moreover, the proposed method enabled our SONE to be applied to a greater variety of tasks.

We developed a learning system for autonomous robots that allows for autonomous exploration of the effective output, and has simple external parameters and a low calculation cost. We propose the concept of self-organizing network elements (SONE) for creating learning systems with these characteristics. We created and evaluated a self-organizing logic circuit by using this concept. Our results indicated this learning system had the characteristics