Anyang Normal University

Optimization of a training program for an athlete is a difficult issue in sport science. It is a delicate trade-off between stimulating performance enhancement and allowing adequate recovery to prevent injury. The paper, presents a detailed structure of the offline optimization of training loads using the DQN architecture. The framework overcomes the simulation gap by applying a data-driven transition model as a digital twin, which can be used to find the optimal training policies without the ethical or safety concerns of conducting the experiment in real-time on athletes. This intelligent model, founded on comprehensive physiological and performance data collected from 25 athletes over a whole training season, has the capacity to dynamically provide ideal training prescriptions like increased intensity, increased volume, or active recovery. Considered data include parameters such as Heart Rate Variability (HRV), sleep quality, training loads, Acute to Chronic Workload Ratio (ACWR), and weekly performance. The proposed architecture uses a feedforward neural network as an estimator of the Q-value function. By optimizing the adaptive epsilon-Greedy policy and the Experience Replay Buffer, the stability and efficiency of the learning process are ensured. In addition, a dual reward function including performance reward and physiological state reward is designed. This function guides the agent towards policies that simultaneously lead to short-term performance improvement and long-term health maintenance of the athlete. The experimental results show that the model has successfully reduced the error rate and has tended to converge to near zero for the loss function. Also, the proposed method has shown a high ability in managing training load and controlling the risk of injury, in such a way that it has been able to dynamically and with high adaptability reduce the risk and always maintain the performance of athletes within the optimal range.

Abstract As the 15th paper in a series reporting on a large reverberation mapping (RM) campaign of super-Eddington accreting massive black holes (SEAMBHs) in active galactic nuclei (AGNs), we present the results of measurements of the Mg ii lines in 18 SEAMBHs monitored spectroscopically from 2017 to 2024. Among these, the time lags of Mg ii have been successfully determined for eight of the 18 objects, thereby expanding the current Mg ii RM sample, particularly at higher accretion rates. By incorporating measurements of the line widths, we determine the masses of their central supermassive black holes. Based on these new measurements, we update the relation between the Mg ii radius and the monochromatic luminosity at 3000 Å ( R MgII – L 3000 relation), yielding a slope of 0.24 ± 0.03, which is slightly shallower than, yet still consistent with, previously reported values. Similar to the H β lines, the Mg ii time lags in SEAMBHs are shorter than those of AGNs with normal accretion rates at comparable luminosities. The deviation of AGNs from the best-fit R MgII – L 3000 relation shows a strong correlation with the accretion rate, while no significant correlation is found between the deviation and the flux ratio of UV iron to Mg ii .

A simple template method was developed for preparation ultrlong Co 11 (HPO 3 ) 8 (OH) 6 /Co 9 S 8 hierarchical tubular architectures, which show excellent electrochemical performance for supercapacitors.

Alzheimer's disease (AD) is a common neurodegenerative disorder among the elderly population. There are currently no effective therapeutic drugs available, the multi-target-directed ligands (MTDLs) strategy has been considered as the promising approach. Given the structural diversity of natural products, Rivastigmine's pharmacophore was integrated with diverse natural product scaffolds to construct a combinatorial compound library. This library was subsequently screened and optimized to identify a novel butyrylcholinesterase (BuChE) inhibitor, compound 3c . The results showed that compound 3c exhibited favorable BuChE inhibitory activity (half-maximal inhibitory concentration (IC 50 ) = 0.43 µmol/L), potential anti-inflammatory potency, good A β 1–42 aggregation inhibitory capacity and remarkable neuroprotective effects. The in vivo study exhibited that 3c significantly ameliorated AlCl 3 -induced zebrafish AD model and scopolamine-induced memory impairment. Collectively, compound 3c was the artificial intelligence (AI)-driven promising multifunctional agent with BuChE inhibition for the treatment of AD. Compound 3c , identified through artificial intelligence (AI)-driven design, showed multifunctional activity against AD, including BuChE inhibition, neuroprotection, and memory improvement in models.

Electrocatalytic oxidation of biomass molecules such as ethanol in hybrid alkaline water electrolysis is more thermodynamically favorable and techno-economic attractive to replace conventional pure water electrooxidation to produce green hydrogen. Herein, the flexible and binder-free hollow microtube catalyst arrays of CoS2/CC, Ni0.04Co0.96S2/CC, Fe0.07Ni0.04Co0.89S2/CC, and Fe0.08Ni0.10Co0.82S2/CC were derived from metal-organic framework arrays anchored on carbon cloth (CC). These arrays exhibited favorable performance in electrochemical water oxidation, ethanol oxidation, and hydrogen evolution processes, because of their obvious advantages in high conductivity and long-term stability. The optimized self-supporting Fe0.08Ni0.10 Co0.82S2/CC electrode composed of a 3D hollow porous microtube structure only needs a low potential of 1.412 and 1.267V (vs. RHE) to deliver 10mA cm-2 current density for alkaline water and ethanol oxidation reactions, respectively. Simultaneously, the pure CoS2/CC electrode presents excellent alkaline hydrogen production property among the series self-supporting electrodes with a low overpotential of 188mV at 10mA cm-2. In this work, it is proved that the hybrid water splitting system, using Fe0.08Ni0.10Co0.82S2/CC and CoS2/CC as anode and cathode, respectively, can effectively reduce the cell voltage to 1.479V to deliver 10mA cm-2 with high pure hydrogen generation and high valued potassium acetate generation.

Two experiments explored a previous finding that 120 repetitions of the same six-item array for change detection resulted in no or very slow learning. This contrasts with learning from six repetitions tested by recall. Shimi and Logie (Quarterly Journal of Experimental Psychology, 72, 1387-1400, 2019) proposed that array repetition for change detection is supported by a limited capacity, temporary visual cache memory, with contents replaced by the next study array, even if it is identical. This is coupled with a weak episodic trace that strengthens across repeated presentations but requires more than 60 repetitions for learning. Experiment 1 tested whether (a) a short study-test interval would result in reliance on the visual cache with no evidence of learning across 120 repetitions, and (b) a longer study-test interval would gradually strengthen the episodic trace but require many repetitions for learning. A 500-ms study-test interval showed no learning after 120 repetitions, and participants reported being unaware of the repetition. A 5,000-ms study-test interval showed performance improvements, but only after 40 repetitions, and participants reported becoming aware of the repetition. In Experiment 2, different arrays on each of 120 trials with short and long study-test intervals showed the same lack of learning found for the 500-ms study-test interval in Experiment 1. Results appear consistent with a limited capacity visual cache memory for change detection that retains the array only for the current trial, working in tandem with a weak episodic trace that accumulates across trials but only supports performance after multiple repetitions and longer study-test intervals.