Wireless sensor networks (WSNs) are typically used with dynamic conditions of task-related environments for sensing(monitoring) and gathering of raw sensor data for subsequent forwarding to a base station. In order to deploy WSNs in real environments, a variety of technical challenges must be addressed. With traditional techniques developed for a specific task, it is hard to react in dynamic situations beyond the scope of the intended task. As a solution to this problem, machine learning (ML) techniques that are able to handle dynamic situations with successful learning process have been applied lately in WSNs. Particularly, deep learning (DL) techniques, a class of ML techniques characterized by the use of deep neural network, are used for WSNs to extract higher level features from raw sensor data. A range of benefits obtained from ML techniques applied to WSNs can be described as reduced computational complexity, increased feasibility in finding optimal solutions, increased energy efficiency, etc. On the other hand, it is found from our survey that large training time and large dataset to get acceptable performance are accompanied with large energy consumption which is not favorable for resource-restrained WSNs. Reviews on the applications of ML techniques in WSNs appeared in the literature. However, few reviews have dealt with the applications of DL techniques in WSNs. In this review, recent developments of ML techniques for WSNs are presented with much emphasis on DL techniques. The DL techniques developed for various applications in WSNs are addressed together with their respective deep neural network architectures.

Antibodies play a crucial role in our immune system. Their ability to bind to and neutralize pathogens opens opportunities to develop antibodies for therapeutic and diagnostic use. Computational methods capable of designing antibodies for a target antigen can revolutionize drug discovery, reducing the time and cost required for drug development. Artificial intelligence (AI) methods have recently achieved remarkable advancements in the design of protein sequences and structures, including the ability to generate scaffolds for a given motif and binders for a specific target. These generative methods have been applied to antigen-conditioned antibody design, with experimental binding confirmed for de novo-designed antibodies. This review surveys current AI methods used in antibody development, focusing on those for antigen-conditioned antibody design. The results obtained by AI-based methodologies in antibody and protein research suggest a promising direction for generating de novo binders for various target antigens.

Games have been used as excellent testbeds for research on artificial intelligence (AI) and computational intelligence for their diversity and complexity. In this article, we present AI World Cup, a set of AI competitions based on the game of soccer. We provide an introduction to the three challenges that concern a robot soccer match using both value-based and image-based state representations. AI Soccer runs the robot soccer match by participants managing each team of five two-wheeled robots. AI Commentator and AI Reporter observe the AI Soccer match and output real-time commentary and a summary article, respectively. Also, we introduce the AI World Cup platform along with rationale behind notable design choices. The official international AI World Cups held in 2018 and 2019 and the AI Masters competition held in 2019 as a part of the World Cyber Games are briefly discussed. Technical aspects of the strategies developed by participants are also discussed.

Optimization of power management of nanogrid based on short-term prediction of PV power production and consequent EV charging/discharging is proposed. Goal of power management is to reduce time-based electricity cost and total delay. To achieve the goal, efficiency in the combined use of PV power and EV charging/discharging power is important. Unlike the PV power used ahead of costly grid power and entirely dependent on weather condition, timing of EV charging/discharging depends on power management scheme. In order to find out the timing for EV charging/discharging, short-term prediction of PV power production is considered as a key contributor. When PV power production is predicted to decrease in short-term, e.g., 10minutes, discharging power of EVs can compensate the loss and, when predicted to increase in short-term, EVs are charged to capitalize on the gain. Short-term prediction of PV power production is performed by long short-term memory (LSTM) network trained and validated by dataset of PV power production over 1 year. In addition, variation of outdoor temperature in relation to indoor temperature is factored in to determine the timing for EV charging/discharging. Our work is comprehensive in that various electric appliances as well as PV source and EVs are taken into account for power management of nanogrid. Simulation results show the cost benefit obtained from the short-term prediction of PV power production and consequent EV charging/discharging while managing peak demand below maximum allowed level.

In multigoal reinforcement learning, an agent interacts with an environment and learns to achieve multiple goals. The goal-conditioned policy is trained to effectively generalize its behavior for multiple goals. During training, the experiences collected by the agent are randomly sampled from a replay buffer. Because biased sampling of achieved goals affects the success rate of a given task, it should be avoided by considering the valid goal space, introduced here as the set of goals to achieve, and the current competence of the policy. To this end, a novel prioritization method for creation of batches, e.g., collections of samples, is proposed. Candidate batches are sampled and associated with costs; in each iteration the batch with the minimum cost is chosen to train the policy. The cost function is modeled by an intended goal, which is proposed as a hypothetical goal that the policy is trying to learn in each cycle, and the information of the valid goal space. The minimum cost of the batch selected for each iteration decreases throughout training as the policy learns to achieve goals near the center of the valid goal space. The proposed batch prioritization method is combined with hindsight experience replay (HER) for experiments in robotic control tasks presented in the OpenAI gym suite to demonstrate learning performance comparable to that of other state-of-the-art prioritization methods. As a result, the proposed batch prioritization method can achieve improved learning performance in 4 out of 5 tasks, particularly for harder tasks. The experimental results suggest that the proposed method for the creation of training batches, using the valid goal space information and current competence of the policy, can enhance learning performance in multigoal tasks with high-dimensional goal space.

Lately, pose estimation based on learning-based Visual Odometry (VO) methods, where raw image data are provided as the input of a neural network to get 6 Degrees of Freedom (DoF) information, has been intensively investigated. Despite its recent advances, learning-based VO methods still perform worse than the classical VO that consists of feature-based VO methods and direct VO methods. In this paper, a new pose estimation method with the help of a Gated Recurrent Unit (GRU) network trained by pose data acquired by an accurate sensor is proposed. The historical trajectory data of the yaw angle are provided to the GRU network to get a yaw angle at the current timestep. The proposed method can be easily combined with other VO methods to enhance the overall performance via an ensemble of predicted results. Pose estimation using the proposed method is especially advantageous in the cornering section which often introduces an estimation error. The performance is improved by reconstructing the rotation matrix using a yaw angle that is the fusion of the yaw angles estimated from the proposed GRU network and other VO methods. The KITTI dataset is utilized to train the network. On average, regarding the KITTI sequences, performance is improved as much as 1.426% in terms of translation error and 0.805 deg/100 m in terms of rotation error.