SmartRolling: A human–machine interface for wheelchair control using EEG and smart sensing techniques     

《Information processing & management》2023,60(3):103262

Smart wheelchairs based on brain–computer interface (BCI) have been widely utilized recently to address certain mobility problems for people with disability. In this paper, we present SmartRolling, an intuitive human–machine interaction approach for the direct control of robotic wheelchair that jointly leverages EEG signals and motion sensing techniques. Specifically, SmartRolling offers two wheelchair-actuation modes for users with different physical conditions: (1) head motion only — people who are severely disabled but able to do basic tasks using eyes and head, and (2) head and hands motion — in addition to type 1, people who can use functioning hands/arms for extra tasks. The system issues operation commands by recognizing different EEG patterns elicited by motor execution (ME) tasks including eye blink, jaw clench, and fist open/close, while at the same time estimates users’ steering intentions based on their facing direction by leveraging inertial measurements and computer vision techniques. The experiment results demonstrate that the proposed system is robust and effective to meets the individual’s needs and has great potential to promote better health.  相似文献   

Online portfolio management via deep reinforcement learning with high-frequency data     

《Information processing & management》2023,60(3):103247

Recently, models that based on Transformer (Vaswani et al., 2017) have yielded superior results in many sequence modeling tasks. The ability of Transformer to capture long-range dependencies and interactions makes it possible to apply it in the field of portfolio management (PM). However, the built-in quadratic complexity of the Transformer prevents its direct application to the PM task. To solve this problem, in this paper, we propose a deep reinforcement learning-based PM framework called LSRE-CAAN, with two important components: a long sequence representations extractor and a cross-asset attention network. Direct Policy Gradient is used to solve the sequential decision problem in the PM process. We conduct numerical experiments in three aspects using four different cryptocurrency datasets, and the empirical results show that our framework is more effective than both traditional and state-of-the-art (SOTA) online portfolio strategies, achieving a 6x return on the best dataset. In terms of risk metrics, our framework has an average volatility risk of 0.46 and an average maximum drawdown risk of 0.27 across the four datasets, both of which are lower than the vast majority of SOTA strategies. In addition, while the vast majority of SOTA strategies maintain a poor turnover rate of approximately greater than 50% on average, our framework enjoys a relatively low turnover rate on all datasets, efficiency analysis illustrates that our framework no longer has the quadratic dependency limitation.  相似文献   

Positivity and quality of college life of business students: The mediating role of learning approaches     

《Studies in Educational Evaluation》2020

Drawing upon the causal agency theory, this study examined the direct and indirect (via learning approaches) links between positivity and quality of college life (QCL) of business students in a transitioning market, Vietnam. Two studies were conducted to test these relationships using structural equation modeling. Findings from study 1, based on a sample of 428 business students, showed that positivity was positively related to QCL. Further, deep learning approaches mediated the relationship between positivity and QCL. Findings from study 2, based on a sample of 416 business students, reconfirmed the role of positivity in QCL. However, the indirect relationship between positivity and QCL was not significant, demonstrating that surface learning approaches did not mediate the relationship between positivity and QCL. The findings contribute to the literature on positivity in higher education by confirming the overarching role that positivity plays, directly and indirectly (via deep learning approaches), in QCL.  相似文献   

Understanding the overvaluation of facial trustworthiness in Airbnb host images     

《International Journal of Information Management》2021

Renting a property via a peer-to-peer platform involves a variety of risks. Humans inherently, subconsciously use facial cues as important shortcuts in making assessments about other persons. On property sharing platforms, such as Airbnb, facial cues can be used in a similar fashion alongside reputational information. According to Dangerous Decisions Theory (DDT), intuitive evaluations of trustworthiness based on faces can bias subsequent assessment of an individual, requiring further information sources to make a more balanced assessment. In this study we apply DDT to demonstrate that evaluations based on perceived facial trustworthiness are overvalued; when combined with reputational measures, such as ‘super host’ status, such assessments are diminished. The study is based on deep learning to classify host faces for a large data set of online accommodation (n = 78,386). The research demonstrates that facial trust cues in online platforms should be treated with caution and must be combined with more objective measures of reputation in order to reduce the effects of overvaluation. The paper concludes with implications for practice and future research.  相似文献   

重视数学阅读,打造深度学习课堂     

李乖乖 《天津教育》2021,(6):183-184

许多小学数学教师已然认识到了开展小学数学阅读活动和制定科学的教学方法,对提高学生的计算能力和思维能力的重要作用。所以,教师在教学过程中要积极探索科学的教学方法培养学生的有效阅读能力,这样学生才能够更好地理解和学习数学知识,提高学生的解题思维能力和学习有效性。以下便对数学阅读活动的开展和深度学习课堂的构建进行分析和探究。  相似文献   

基于动量算法的常数模新算法     

王小珍  饶伟  王奇珍  郭业才 《中国科技信息》2006,(14):260-261,290

针对传统常数模算法收敛速度慢的缺点，提出了一种基于动量算法的常数模算法。该算法的权向量更新公式中添加了动量因子，以提高收敛速度。利用水声信道对算法的性能进行仿真研究，仿真结果表明：与常数模算法相比，该算法能较快收敛。  相似文献   

一种有效的混合均衡算法     

马汝星  唐金花  陈偕雄  朱小富 《科技通报》2004,20(3):229-233

提出一种基于自适应均衡算法和盲均衡算法的混合算法DDLMS-Godard算法，该算法充分发挥了DDLMS算法和Godard算法的优势，克服了各自的缺陷，在MATLAB6．0平台上的性能仿真结果表明，DDLMS-Godard算法不仅收敛速率快，而且稳定性好，具有很好的实用价值。  相似文献   

Research on evaluation model for interactive classroom enabled with mobile terminals     

Yushun Li  Qun Lin  Chunlan Qian 《Interactive Learning Environments》2019,27(2):163-180

Information and Communication Technology (ICT) has improved education widely in China, transforming traditional teaching into an interactive one. It is important to build a new evaluation model to measure teaching efficiency in an “ICT-enabled classroom”. This study designed an evaluation model named “TPOCME Deep classroom” through ongoing five iterations. It includes six dimensions, which are higher-order Thinking, classroom Participation, Openness of educational system, Cooperative learning, Meaningful learning and Effectiveness of technology use (named TPOCME). This model helps educational researchers and teachers gain a comprehensive understanding of ICT in education.  相似文献   

NASA深空网的改造计划及进展     

王天祥 《环球赛鸽科技》2015,(19)

详细介绍了美国的深空网增强计划及进展。  相似文献   

相位谱拟合车载自组网中功率增益优化方法     

金鑫 《科技通报》2015,(2):79-81

车载自组网WSN无线传感网络的接收/发射功率的无偏均衡设计,可以提高网络节点寿命,保证系统的稳定性,提高车载自组网的通信能力和网络连接率。传统方法采用的固定步长加权的滤波设计算法进行功率均衡设计,导致车载自组网的相邻节点功率均衡偏差,提出一种基于无偏均衡和相位谱拟合的车载自组网中功率增益优化算法,进行无偏均衡的车载自组网中网络连接率优化设计,确定节点一跳范围内的邻居节点,实现功率谱的估计,在功率均衡设计中,进行可变步长滤波设计,反映了信道幅值、时间和频率之间关系,对相位谱做最小二乘拟合求斜率,实现WSN可变步长滤波接收/发射功率无偏均衡算法改进。研究结果表明,该算法实现了接收/发射功率的无偏均衡,网传输距离提高,具有较大的功率增益,网络连接率较传统方法提高,展示了优越的应用性能。  相似文献