Traffic Signal Timing via Deep Reinforcement Learning

In this paper, we propose a set of algorithms to design signal timing plans via deep reinforcement learning. The core idea of this approach is to set up a deep neural network (DNN) to learn the Q-function of reinforcement learning from the sampled traffic state/control inputs and the corresponding traffic system performance output. Based on the obtained DNN, we can find the appropriate signal timing policies by implicitly modeling the control actions and the change of system states. We explain the possible benefits and implementation tricks of this new approach. The relationships between this new approach and some existing approaches are also carefully discussed.      

Control 5.0: From Newton to Merton in Popper's Cyber-Social-Physical Spaces

The future of control in cyberspace of parallel worlds is discussed. It argues for the coming age of Control 5.0, the control technology for the new IT capable of dealing with artificial worlds with VR, AR, AI and robotics. The discipline of automation needs a new interpretation of its core knowledge and skill set of modeling, analysis, and control for cyber-socialphysical systems, and a paradigm shift from Newtonian Systems with Newton's Laws or Big Laws with Small Data to Mertonian Systems with Merton's Laws or Small Laws with Big Data.      

Ubiquitous Mining with Interactive Data Mining Agents

Due to the increasing availability and sophistication of data recording techniques, multiple information sources and distributed computing are becoming the important trends of modern information systems. Many applications such as security informatics and social computing require a ubiquitous data analysis platform so that decisions can be made rapidly under distributed and dynamic system environments. Although data mining has now been popularly used to achieve such goals, building a data mining system is, however, a nontrivial task, which may require a complete understanding on numerous data mining techniques as well as solid programming skills. Employing agent techniques for data analysis thus becomes increasingly important, especially for users not familiar with engineering and computational sciences, to implement an effective ubiquitous mining platform. Such data mining agents should, in practice, be intelligent, complete, and compact. In this paper, we present an interactive data mining agent — OIDM (online interactive data mining), which provides three categories (classification, association analysis, and clustering) of data mining tools, and interacts with the user to facilitate the mining process. The interactive mining is accomplished through interviewing the user about the data mining task to gain efficient and intelligent data mining control. OIDM can help users find appropriate mining algorithms, refine and compare the mining process, and finally achieve the best mining results. Such interactive data mining agent techniques provide alternative solutions to rapidly deploy data mining techniques to broader areas of data intelligence and knowledge informatics.     

The coming of CSP: automated assembly processes for high-volume chip scale packaging

Chip scale packaging (CSP) is emerging as a major player within the semiconductor packaging market. Processing in the back-end requires optimizing the use of intermediate carrier configurations such as tape-in-carrier and strip-based substrates. This optimization of carrier configurations must occur in all processes related to CSP assembly, including die attach, wire bond, encapsulation, ball attach, laser mark, singulate, inspect, and test. CSP is considered an SMT and, as such, can be processed using existing lines. However, CSP is different from SMT and has its own assembly problems. One of the critical problems is that, with thinner package profile requirement, thinner substrate is necessary. Another important problem that should be considered is the failure of solder joints because of the coefficient of thermal expansion mismatch among the substrate, lead packages, and the board when the thermal load is severe. To address this point, several studies using moire interferometry have been conducted to measure thermal deformations that accumulate during thermal cycles.     

Artificial societies for integrated and sustainable development of metropolitan systems

Unfortunately, we can't solve transportation problems by focusing on transportation systems alone. Rather, we must consider the combined effects with other metropolitan systems. Artificial systems, based on artificial societies and agent-modeling technology, are effective tools for this purpose. Metropolitan transportation, logistics, and ecosystems are intrinsically open, dynamic, unpredictable, and complex in their behaviors and effects. We must adopt a management and control strategy for those systems based on continuous investigation and improvement and should use computational experiments with artificial systems to overcome the difficulty of experimenting with real systems.     

平行皮肤:基于视觉的皮肤病分析框架

随着计算机与人工智能的快速发展,基于图像感知的皮肤病分析方法取得一些成果.然而,以深度学习为主的计算机辅助分析方法依赖于领域专家标注的医学大数据,诊断结果缺乏医学可解释性.为此,文中提出基于视觉的皮肤病分析统一框架——平行皮肤.启发于ACP方法与平行医学图像分析框架,通过构建人工皮肤图像系统实现数据选择与生成,通过预测学习的计算实验完成诊断分析模型构建与评估,并利用描述学习与指示学习融合专家知识,引导人工图像系统数据生成与选择,从而实现闭环诊断分析模型优化.     

DAO to HANOI via DeSci: AI Paradigm Shifts from AlphaGo to ChatGPT

From AlphaGo to ChatGPT,the field of AI has launched a series of remarkable achievements in recent years.Analyzing,comparing,and summarizing these achievements at the paradigm level is important for future AI innovation,but has not received sufficient attention.In this paper,we give an overview and perspective on machine learning paradigms.First,we propose a paradigm taxonomy with three levels and seven dimensions from a knowledge perspective.Accordingly,we give an overview on three basic and tw...     

Smart Cars on Smart Roads: An IEEE Intelligent Transportation Systems Society Update

To promote tighter collaboration between the IEEE Intelligent Transportation Systems Society and the pervasive computing research community, the authors introduce the ITS Society and present several pervasive computing-related research topics that ITS Society researchers are working on. This department is part of a special issue on Intelligent Transportation.     

A robotic vision system for object identification and manipulation using synergetic pattern recognition

This paper discusses the application of the synergetic pattern recognition method to a robotic vision system for workpiece identification and manipulation in automated flexible manufacturing environments. The original synergetic algorithm is extended to allow its pattern attention parameters to have different values. Stability analysis of the extended recognition model indicates that the prototype patterns are the only stable patterns and undesired spurious patterns cannot exist. A simple scheme for tuning attention parameters is developed. Simulation results show that the number of object misclassification is reduced significantly with this extension. In addition, an image preprocessing procedure enables synergetic recognition to be simultaneously invariant to spatial pattern translation, rotation, and scaling; while an approach for recovering position, orientation, and size information is also proposed. Simple and efficient task-directed and object-specific strategies for robotic workpiece manipulation are now easy to implement based on these results and procedures.