MRPⅡ is an applied science which incorporates management skill,computer and manufacturing technology.It requires that enterprise trace the environment changes from inside and outside and revise the management strategy of its business according to the principle of supply and demand chain and information collection. MRPⅡ sets purposely a pattern for production system and economic activities and tries to balance the enterprise’s source of production imaterials and requirment of management tasts through the pattern of motion so that the enterprise can surely achieve its goal. MRPⅡ is made to establish a cooporative partnership between suppliers and demanders on the basis of mutual trust,benefit,aid and rewprocity, it embodies the essence of exquisite production and quick manufacturing and aims to enhance the advantageous competition for the enterprise.The wide-spread introduction is demanded by the new pattern of micro-management of modern business management.
This paper studies a new task of federated learning (FL) for semantic parsing, where multiple clients collaboratively train one global model without sharing their semantic parsing data. By leveraging data from multiple clients, the FL paradigm can be especially beneficial for clients that have little training data to develop a data-hungry neural semantic parser on their own. We propose an evaluation setup to study this task, where we re-purpose widely-used single-domain text-to-SQL datasets as clients to form a realistic heterogeneous FL setting and collaboratively train a global model. As standard FL algorithms suffer from the high client heterogeneity in our realistic setup, we further propose a novel LOss Reduction Adjusted Re-weighting (Lorar) mechanism, which adjusts each client’s contribution to the global model update based on its training loss reduction during each round. Our intuition is that the larger the loss reduction, the further away the current global model is from the client’s local optimum, and the larger weight the client should get. By applying Lorar to three widely adopted FL algorithms (FedAvg, FedOPT and FedProx), we observe that their performance can be improved substantially on average (4%-20% absolute gain under MacroAvg) and that clients with smaller datasets enjoy larger performance gains. In addition, the global model converges faster for almost all the clients.
Question answering over knowledge bases is considered a difficult problem due to the challenge of generalizing to a wide variety of possible natural language questions. Additionally, the heterogeneity of knowledge base schema items between different knowledge bases often necessitates specialized training for different knowledge base question-answering (KBQA) datasets. To handle questions over diverse KBQA datasets with a unified training-free framework, we propose KB-BINDER, which for the first time enables few-shot in-context learning over KBQA tasks. Firstly, KB-BINDER leverages large language models like Codex to generate logical forms as the draft for a specific question by imitating a few demonstrations. Secondly, KB-BINDER grounds on the knowledge base to bind the generated draft to an executable one with BM25 score matching. The experimental results on four public heterogeneous KBQA datasets show that KB-BINDER can achieve a strong performance with only a few in-context demonstrations. Especially on GraphQA and 3-hop MetaQA, KB-BINDER can even outperform the state-of-the-art trained models. On GrailQA and WebQSP, our model is also on par with other fully-trained models. We believe KB-BINDER can serve as an important baseline for future research. Our code is available at https://github.com/ltl3A87/KB-BINDER.
Purpose To develop a predictive model for fall risk in pre-frail older adults, providing a basis for early identification and prevention of falls among this population. Method This was a multicenter prospective cohort study. A total of 473 pre-frail older adults were included, 335 as the training set and 142 as the test set. Univariate and stepwise binary logistic regression analyses were conducted to identify the relationship between pre-frail and fall risk and establish the frailty risk prediction nomogram. The nomogram was constructed based on the results of logistic regression. The model assessment relied on the receiver operating characteristic (ROC) curve, Hosmer–Lemeshow test, calibration curves, and decision curve analysis. Results Fall incidence rate among pre-frail older adults within 6 months was 13.63%. The final fall risk prediction model identified that sex, history of falls in the past year, visual impairment, increased nocturia, and fear of falling are the most critical risk factors for falls in pre-frail older adults. The model exhibited good accuracy in the testing set, with an area under the ROC curve of 0.825 (95% confidence interval [0.736, 0.914]). Conclusion Pre-frail older adults have a higher incidence of falls. The logistic regression model constructed in this study shows promising predictive capabilities and can be used as a screening tool to identify pre-frail older adults at high risk of falls in clinical practice. We anticipate that this model will assist clinical nurses in enhancing the efficiency of fall prevention efforts and reducing the incidence of falls among pre-frail older adults. [ Research in Gerontological Nursing, 18 (1), 29–39.]
Background: Unhealthy lifestyles among adolescents are reaching alarming levels and have become a major public health problem. This study aimed to assess the relationship between sleep time, physical activity (PA) time, screen time (ST), and nutritional literacy (NL). Methods: This cross-sectional online study involving adolescents aged 10–18 years was conducted in September 2020 in 239 schools in Chongqing, China. NL was measured using the “Nutrition Literacy Scale for middle school students in Chongqing (CM-NLS)”. According to the recommended by the Chinese dietary guidelines (2022), we divided the sleep time of junior high school students into <9 h and ≥9 h, high school students into <8 h and ≥8 h, divided the workdays into weekend PA time < 1 h and ≥1 h, and divided the workdays into weekend ST < 2 h and ≥2 h. The multinomial logistic regression model was used to examine the association. Results: A total of 18,660 adolescents (50.2% males) were included. The proportion of participants that were junior high school students and attended boarding schools was 57.2% and 65.3%, respectively. Compared with senior high school students, junior high school students had a higher level of NL. Whether on workdays or weekends, participants with sleep time ≥ 8/9 h, PA time ≥ 1 h, and ST < 2 h per day had higher levels of NL. On weekdays, participants who met the sleep time ≥ 8 h/9 h (OR = 1.48, 95% CI: 1.36, 1.62) and PA time ≥ 1 h (OR = 1.69, 95% CI: 1.59, 1.81) had higher reporting of NL levels. Conclusions: Sleep time, PA time, and ST were positively correlated with NL among adolescents, especially junior high school students.
Acupuncture can improve explosive force production and affect joint stiffness by affecting muscle activation levels. This study aims to explore the effects of true acupuncture (TA) compared with sham acupuncture (SA) on the explosive force production and stiffness of the knee joint in healthy male subjects. Twenty subjects were randomly divided into the TA group (n = 10) and SA group (n = 10) to complete isokinetic movement of the right knee joint at a speed of 240°/s before and after acupuncture. Futu (ST32), Liangqiu (ST34), Zusanli (ST36), Xuehai (SP10), and Chengshan (BL57) were selected for acupuncture. The intervention of SA is that needles with a blunt tip were pushed against the skin, giving an illusion of insertion. The results showed that acupuncture and the intervention time had a significant interaction effect on knee joint explosive force and joint stiffness (p < 0.05). The average maximum (max) torque, average work, average power, average peak power and total work of the TA group increased significantly after acupuncture (p < 0.05), while the SA group did not (p > 0.05). Therefore, true acupuncture can immediately improve the explosive force and joint stiffness of the male knee joint by inducing post-activation potentiation (PAP) and/or De-Qi.
Background: Gait asymmetry is often accompanied by the bilateral asymmetry of the lower limbs. The transcranial direct current stimulation (tDCS) technique is widely used in different populations and scenarios as a potential tool to improve lower limb postural control.Research question: Whether cerebral cortex bilateral tDCS has an intervention effect on postural control as well as bilateral symmetry when crossing obstacles in healthy individuals?Method: Twenty healthy females were recruited in this prospective study. Each participant walked and crossed a height-adjustable obstacle. Two-way repeated ANOVA was used to evaluate the effect of group (tDCS and sham-tDCS) and height (30%, 20%, and 10% leg length) on the spatiotemporal and maximum joint angle parameters for lower limb crossing obstacles. The Bonferroni posthoc test and paired t-test were used to determine the significance of the interaction effect or main effect. The statistically significant differences were set at p<0.05.Results: The Swing time (SW) gait asymmetry (GA), Stance time (ST) GA, SW/ST GA, and Leading limb hip-knee-ankle maximum joint angles decreased in the tDCS group compared to the sham-tDCS group at 30%, 20% LL crossing height (P<0.05), whereas there was no difference between the tDCS group and the sham tDCS group at 10% LL crossing height (P>0.05). Trailing limb hip-knee-ankle maximum joint angles were unchanged in the tDCS group compared to the sham-tDCS group at 30%, 20%, and 10% LL crossing height (P>0.05).Significance: The decreased SW, ST and SW/ST GA and leading/trailing limb joint angles changes at higher crossing heights indicated that tDCS helped to reduce bilateral asymmetry, simultaneously, tDCS has a certain effect on the agility, end-point precision control, and postural stability of the lower limb when crossing obstacles for healthy females.
This study aims to explore the effect of real-time visual feedback (VF) information of the pres-sure of center (COP) provided by intelligent insoles on balance training in a one leg stance (OLS) and tandem stance (TS) posture. Thirty healthy female college students were randomly assigned to the visual feedback balance training group (VFT), non-visual feedback balance training group (NVFT), and control group (CG). The balance training includes: OLS, tandem Stance (dominant leg behind, TSDL), tandem stance (non-dominant leg behind, TSNDL). The training lasted 4 weeks, the training lasts 30 minutes at an interval of 1 days. There was a sig-nificant difference in the interaction effect between Groups*Times of the COP parameters (p<0.05) for OLS. There was no significant difference in the interaction effect between Groups*Times of the COP parameters (p>0.05) for TS. The main effect of the COP parameters was a significant difference in Times (p<0.05). The COP displacement, velocity, radius, and area in VFT significantly decreased after training (p < 0.05). Therefore, the visual feedback technology of intelligent auxiliary equipment during balance training can enhance the benefit of training. The use of smart wearable devices in OLS balance training may improve the visual and physical balance integration ability.
With the rapid development of drones, unmanned vehicles and robotics industries, VLAM has become a hot technology. In particular, the birth of 5G-powered UAV has promoted the emergence of more industrial applications, making it the most core and indispensable role in many scenarios. The loop closure detection can decrease the accumulative total of error during the process of VSLAM. Former loop closure detection methods always rely on artificially features, which are not robust, making it hard to deal with changing complex scenarios. The later deep learning-based methods are considered to be better solutions for loop closure detection. However, due to the simple network structure, there is still a lot of room for improvement. This paper proposes a more complex neural network to achieve loop closure detection. This approach adopts a fish-shaped deep neural network backbone, which can extract and fuse data features at different levels. Experiments demonstrate the feasibility and effectiveness of this backbone in loop closure detection problems.
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)