Reinforcement learning has gathered much attention in recent years due to its rapid development and rich applications, especially on control systems and robotics. When tackling real-world applications with reinforcement learning method, the corresponded Markov decision process may have huge discrete or even continuous state/action space. Deep reinforcement learning has been studied for handling these issues through deep learning for years, and one promising branch is the actor-critic architecture. Many past studies leveraged multiple critics to enhance the accuracy of evaluation of a policy for addressing the overestimation and underestimation issues. However, few studies have considered the architecture with multiple actors together with multiple critics. This study proposes a novel multi-actor multi-critic (MAMC) deep deterministic reinforcement learning method. The proposed method has three main features, including selection of actors based on non-dominated sorting for exploration with respect to skill and creativity factors, evaluation for actors and critics using a quantile-based ensemble strategy, and exploiting actors with best skill factor. Theoretical analysis proves the learning stability and bounded estimation bias for the MAMC. The present study examines the performance on a well-known reinforcement learning benchmark MuJoCo. Experimental results show that the proposed framework outperforms state-of-the-art deep deterministic based reinforcement learning methods. Experimental analysis also indicates the proposed components are effective. Empirical analysis further investigates the validity of the proposed method, and shows its benefit on complicated problems. The source code can be found at this https URL.

We present a study of $B\to X(3872)K$ with X(3872) decaying to $D^{*0}\bar D^0$using a sample of 657 million$B\bar B$ pairs recorded at the $\Upsilon(4S)$ resonance with the Belle detector at the KEKB asymmetric-energy $e^+e^-$ collider. Both $D^{*0}\to D^0\gamma$ and $D^{*0}\to D^0\pi^0$ decay modes are used. We find a peak of $50.1^{+14.8}_{-11.1}$ events with a mass of $(3872.9^{+0.6 +0.4}_{-0.4 -0.5}){\rm MeV}/c^2$, a width of $(3.9^{+2.8 +0.2}_{-1.4 -1.1}){\rm MeV}/c^2$and a product branching fraction${\cal B}(B\to X(3872)K)\times{\cal B}(X(3872)\to D^{*0}\bar D^0)=(0.80\pm0.20\pm0.10)\times10^{-4}$, where the first errors are statistical and the second ones are systematic. The significance of the signal is $6.4\sigma$. The difference between the fitted mass and the $D^{*0}\bar D^0$ threshold is calculated to be $(1.1^{+0.6 +0.1}_{-0.4 -0.3}){\rm MeV}/c^2$. We also obtain an upper limit on the product of branching fractions ${\cal B}(B\to Y(3940)K)\times{\cal B}(Y(3940)\to D^{*0}\bar D^0)$of$0.67\times10^{-4}$ at 90% CL.

The dark photon, $A^\prime$, and the dark Higgs boson, $h^\prime$, are hypothetical constituents featured in a number of recently proposed Dark Sector Models. Assuming prompt decays of both dark particles, we search for their production in the so-called Higgs-strahlung channel, $e^+e^- \rightarrow A^\prime h'$, with $h^\prime \rightarrow A^\prime A^\prime$. We investigate ten exclusive final-states with $A^\prime \rightarrow e^+e^-$, $\mu^+\mu^-$, or $\pi^+\pi^-$, in the mass ranges $0.1$~GeV/$c^2$~< m_{A^\prime} < 3.5~GeV/$c^2$ and $0.2$~GeV/$c^2$~< m_{h'} < 10.5~GeV/$c^2$. We also investigate three inclusive final-states, $2(e^+e^-)X$, $2(\mu^+\mu^-)X$, and $(e^+e^-)(\mu^+\mu^-)X$, where $X$ denotes a dark photon candidate detected via missing mass, in the mass ranges $1.1$~GeV/$c^2$~< m_{A^\prime} < 3.5~GeV/$c^2$ and $2.2$~GeV/$c^2$~< m_{h'} < 10.5~GeV/$c^2$. Using the entire $977\,\mathrm{fb}^{-1}$ data set collected by Belle, we observe no significant signal. We obtain individual and combined 90$\%$ confidence level upper limits on the branching fraction times the Born cross section, $\cal B \times \sigma_{\mathrm{Born}}$, on the Born cross section, $\sigma_{\mathrm{Born}}$, and on the dark photon coupling to the dark Higgs boson times the kinetic mixing between the Standard Model photon and the dark photon, $\alpha_D \times \epsilon^2$. These limits improve upon and cover wider mass ranges than previous experiments. The limits from the final-states $3(\pi^+\pi^-)$ and $2(e^+e^-)X$ are the first placed by any experiment. For $\alpha_D$ equal to 1/137, m_{h'}< 8 GeV/$c^2$, and m_{A^\prime}< 1 GeV/$c^2$, we exclude values of the mixing parameter, $\epsilon$, above $\sim 8 \times 10^{-4}$.

In recent years, neuromorphic computing and spiking neural networks (SNNs) have ad-vanced rapidly through integration with deep learning. However, the performance of SNNs still lags behind that of convolutional neural networks (CNNs), primarily due to the limited information capacity of spike-based data. Although some studies have attempted to improve SNN performance by training them with non-spiking inputs such as static images, this approach deviates from the original intent of neuromorphic computing, which emphasizes spike-based information processing. To address this issue, we propose a Neuron-like Encoding method that generates spike data based on the intrinsic operational principles and functions of biological neurons. This method is further enhanced by the incorporation of an artificial pho-toreceptor layer, enabling spike data to carry both color and luminance information, thereby forming a complete visual spike signal. Experimental results using the Integrate-and-Fire neuron model demonstrate that this biologically inspired approach effectively increases the information content of spike signals and improves SNN performance, all while adhering to neuromorphic principles. We believe this concept holds strong potential for future development and may contribute to overcoming current limitations in neuro-morphic computing, facilitating broader applications of SNNs.

Falls are the public health issue for the elderly all over the world since the fall-induced injuries are associated with a large amount of healthcare cost. Falls can cause serious injuries, even leading to death if the elderly suffers a "long-lie". Hence, a reliable fall detection (FD) system is required to provide an emergency alarm for first aid. Due to the advances in wearable device technology and artificial intelligence, some fall detection systems have been developed using machine learning and deep learning methods to analyze the signal collected from accelerometer and gyroscopes. In order to achieve better fall detection performance, an ensemble model that combines a coarse-fine convolutional neural network and gated recurrent unit is proposed in this study. The parallel structure design used in this model restores the different grains of spatial characteristics and capture temporal dependencies for feature representation. This study applies the FallAllD public dataset to validate the reliability of the proposed model, which achieves a recall, precision, and F-score of 92.54%, 96.13%, and 94.26%, respectively. The results demonstrate the reliability of the proposed ensemble model in discriminating falls from daily living activities and its superior performance compared to the state-of-the-art convolutional neural network long short-term memory (CNN-LSTM) for FD.

Modeling large dependent datasets in modern time series analysis is a crucial research area. One effective approach to handle such datasets is to transform the observations into density functions and apply statistical methods for further analysis. Income distribution forecasting, a common application scenario, benefits from predicting density functions as it accounts for uncertainty around point estimates, leading to more informed policy formulation. However, predictive modeling becomes challenging when dealing with mixed-frequency data. To address this challenge, this paper introduces a mixed data sampling regression model for probability density functions (PDF-MIDAS). To mitigate variance inflation caused by high-frequency prediction variables, we utilize exponential Almon polynomials with fewer parameters to regularize the coefficient structure. Additionally, we propose an iterative estimation method based on quadratic programming and the BFGS algorithm. Simulation analyses demonstrate that as the sample size for estimating density functions and observation length increase, the estimator approaches the true value. Real data analysis reveals that compared to single-sequence prediction models, PDF-MIDAS incorporating high-frequency exogenous variables offers a wider range of application scenarios with superior fitting and prediction performance.

Background: Fluctuating hearing loss is characteristic of Meniere's Disease (MD) during acute episodes. However, no reliable audiometric hallmarks are available for counselling the hearing recovery possibility. Aims/Objectives: To find parameters for predicting MD hearing outcomes. Material and Methods: We applied machine learning techniques to analyse transient-evoked otoacoustic emission (TEOAE) signals recorded from patients with MD. Thirty unilateral MD patients were recruited prospectively after onset of acute cochleo-vestibular symptoms. Serial TEOAE and pure-tone audiogram (PTA) data were recorded longitudinally. Denoised TEOAE signals were projected onto the three most prominent principal directions through a linear transformation. Binary classification was performed using a support vector machine (SVM). TEOAE signal parameters, including signal energy and group delay, were compared between improved and nonimproved groups using Welchs t-test. Results: Signal energy did not differ (p = 0.64) but a significant difference in 1-kHz (p = 0.045) group delay was recorded between improved and nonimproved groups. The SVM achieved a cross-validated accuracy of >80% in predicting hearing outcomes. Conclusions and Significance: This study revealed that baseline TEOAE parameters obtained during acute MD episodes, when processed through machine learning technology, may provide information on outer hair cell function to predict hearing recovery.

We present an end-to-end, interpretable, deep-learning architecture to learn a graph kernel that predicts the outcome of chronic disease drug prescription. This is achieved through a deep metric learning collaborative with a Support Vector Machine objective using a graphical representation of Electronic Health Records. We formulate the predictive model as a binary graph classification problem with an adaptive learned graph kernel through novel cross-global attention node matching between patient graphs, simultaneously computing on multiple graphs without training pair or triplet generation. Results using the Taiwanese National Health Insurance Research Database demonstrate that our approach outperforms current start-of-the-art models both in terms of accuracy and interpretability.

We report the observation of two narrow structures in the mass spectra of the pi+-Y(nS) (n=1,2,3) and pi+-hb(mP)(m$ (m=1,2) pairs that are produced in association with a single charged pion in Y(5S) decays. The measured masses and widths of the two structures averaged over the five final states are M_1=(10607.2+-2.0) MeV/c2, Gamma_1=(18.4+-2.4) MeV and M_2=(10652.2+-1.5) MeV/c2, Gamma_2=(11.5+-2.2) MeV. The results are obtained with a 121.4 1/fb data sample collected with the Belle detector in the vicinity of the Y(5S) resonance at the KEKB asymmetric-energy e+e- collider.

We have searched for the lepton-flavor-violating decays tau- -> ell-K0s and ell-K0sK0s (ell = e or mu), using a data sample of 671 fb^-1 collected with the Belle detector at the KEKB asymmetric-energy e^+e^- collider. No evidence for a signal was found in any of the decay modes, and we set the following upper limits for the branching fractions: B(tau^- -> e^-K0s) < 2.6 x 10^-8, B(tau^- -> \mu^-K0s) < 2.3 x 10^-8, B(tau^- -> e^-K0sK0s) < 7.1 x 10^-8 and B(tau^- -> mu^-K0sK0s) < 8.0 x 10^-8 at the 90% confidence level.

With the full data sample of $772 \times 10^6$ $B{\bar B}$ pairs recorded by the Belle detector at the KEKB electron-positron collider, the decay $\bar{B} \rightarrow D^* \tau^- \bar{\nu}_\tau$ is studied with the hadronic $\tau$ decays $\tau^- \rightarrow \pi^- \nu_\tau$ and $\tau^- \rightarrow \rho^- \nu_\tau$. The $\tau$ polarization $P_\tau(D^*)$ in two-body hadronic $\tau$ decays is measured, as well as the ratio of the branching fractions $R(D^{*}) = \mathcal{B}(\bar {B} \rightarrow D^* \tau^- \bar{\nu}_\tau) / \mathcal{B}(\bar{B} \rightarrow D^* \ell^- \bar{\nu}_\ell)$, where $\ell^-$ denotes an electron or a muon. Our results, $P_\tau(D^*) = -0.38 \pm 0.51 {\rm (stat)} ^{+0.21}_{-0.16} {\rm (syst)}$ and $R(D^*) = 0.270 \pm 0.035{\rm (stat)} ^{+0.028}_{-0.025}{\rm (syst)}$, are consistent with the theoretical predictions of the Standard Model. The polarization values of $P_\tau(D^*) > +0.5$ are excluded at the 90\% confidence level.

The number of independent sets is equivalent to the partition function of the hard-core lattice gas model with nearest-neighbor exclusion and unit activity. We study the number of independent sets $m_{d,b}(n)$ on the generalized Sierpinski gasket $SG_{d,b}(n)$ at stage $n$ with dimension $d$ equal to two, three and four for $b=2$, and layer $b$ equal to three for $d=2$. The upper and lower bounds for the asymptotic growth constant, defined as $z_{SG_{d,b}}=\lim_{v \to \infty} \ln m_{d,b}(n)/v$ where $v$ is the number of vertices, on these Sierpinski gaskets are derived in terms of the results at a certain stage. The numerical values of these $z_{SG_{d,b}}$ are evaluated with more than a hundred significant figures accurate. We also conjecture the upper and lower bounds for the asymptotic growth constant $z_{SG_{d,2}}$ with general $d$.

We search for lepton-flavor-violating and lepton-number-violating tau decays into a lepton (l = electron or muon) and two charged mesons (h, h' = pion or Kaon) using 854 fb^{-1} of data collected with the Belle detector at the KEKB asymmetric-energy e^+e^- collider. We obtain 90% confidence level upper limits on the tau to lhh' branching fractions in the range (2.0-8.4)*10^{-8}. These results improve upon our previously published upper limits by factors of about 1.8 on average.

We observe the decay B0 to p pbar K*0 with a branching fraction of (1.18^{+0.29}_{-0.25} (stat.) \pm 0.11 (syst.)) \times 10^{-6}. The statistical significance is 7.2 sigma for the signal in the low ppbar mass region. We study the decay dynamics of B0 to p pbar K*0 and compare it with B+ to p pbar K*+. The K*0 meson is found to be almost 100% polarized (with a fraction of (101 \pm 13 \pm 3)% in the helicity zero state), while the K*+ meson has a (32 \pm 17 \pm 9)% fraction in the helicity zero state. The direct CP asymmetries for B0 to p pbar K*0 and B+ to p pbar K*+ are measured to be -0.08\pm 0.20\pm 0.02 and -0.01\pm 0.19\pm 0.02, respectively. We also study the characteristics of the low mass ppbar enhancements near threshold and the associated angular distributions. In addition, we report improved measurements of the branching fractions BF(B+ to p pbar K*+) = (3.38^{+0.73}_{-0.60} \pm 0.39) \times 10^{-6} and BF(B0 to p pbar K0) = (2.51^{+0.35}_{-0.29} \pm 0.21) \times 10^{-6}, which supersede our previous measurements. These results are obtained from a 492 fb^{-1} data sample collected near the Upsilon(4S) resonance with the Belle detector at the KEKB asymmetric-energy e^+ e^- collider.

We report a study of the decay $D^0 \to K^0_S K^0_S$ using 921~fb$^{-1}$ of data collected at or near the $\Upsilon(4S)$ and $\Upsilon(5S)$ resonances with the Belle detector at the KEKB asymmetric energy $e^+e^-$ collider. The measured time-integrated $CP$ asymmetry is $ A_{CP}(D^0 \to K^0_S K^0_S) = (-0.02 \pm 1.53 \pm 0.02 \pm 0.17) \%$, and the branching fraction is $\mathcal{B} (D^{0}\rightarrow K_{S}^{0}K_{S}^{0})$ = (1.321 $\pm$ 0.023 $\pm$ 0.036 $\pm$ 0.044) $\times$ 10$^{-4}$, where the first uncertainty is statistical, the second is systematic, and the third is due to the normalization mode ($D^0 \to K_S^0 \pi^0$). These results are significantly more precise than previous measurements available for this mode. The $A_{CP}$ measurement is consistent with the standard model expectation.

Recently, falsified images have been found in papers involved in research misconducts. However, although there have been many image forgery detection methods, none of them was designed for molecular-biological experiment images. In this paper, we proposed a fast blind inquiry method, named FBI$_{GEL}$, for integrity of images obtained from two common sorts of molecular experiments, i.e., western blot (WB) and polymerase chain reaction (PCR). Based on an optimized pseudo-background capable of highlighting local residues, FBI$_{GEL}$ can reveal traceable vestiges suggesting inappropriate local modifications on WB/PCR images. Additionally, because the optimized pseudo-background is derived according to a closed-form solution, FBI$_{GEL}$ is computationally efficient and thus suitable for large scale inquiry tasks for WB/PCR image integrity. We applied FBI$_{GEL}$ on several papers questioned by the public on \textbf{PUBPEER}, and our results show that figures of those papers indeed contain doubtful unnatural patterns.

Multivariate networks are commonly found in real-world data-driven applications. Uncovering and understanding the relations of interest in multivariate networks is not a trivial task. This paper presents a visual analytics workflow for studying multivariate networks to extract associations between different structural and semantic characteristics of the networks (e.g., what are the combinations of attributes largely relating to the density of a social network?). The workflow consists of a neural-network-based learning phase to classify the data based on the chosen input and output attributes, a dimensionality reduction and optimization phase to produce a simplified set of results for examination, and finally an interpreting phase conducted by the user through an interactive visualization interface. A key part of our design is a composite variable construction step that remodels nonlinear features obtained by neural networks into linear features that are intuitive to interpret. We demonstrate the capabilities of this workflow with multiple case studies on networks derived from social media usage and also evaluate the workflow with qualitative feedback from experts.

Using 980 $fb^-1$ of data collected with the Belle detector operating at the KEKB asymmetric-energy e^+e^- collider, we report a study of the electromagnetic decays of excited {charmed baryons} $\Xi_c(2790)$ and $\Xi_c(2815)$. A clear signal (8.6 standard deviations) is observed for $\Xi_c(2815)^0 \to \Xi_c^0\gamma$, and we measure: $B[\Xi_c(2815)^0 \to \Xi_c^0\gamma]/B[\Xi_c(2815)^0 \to \Xi_c(2645)^+\pi^- \to \Xi_c^0\pi^+\pi^-] = 0.41 \pm 0.05 \pm 0.03$. We also present evidence (3.8 standard deviations) for the similar decay of the $\Xi_c(2790)^0$ and measure: $B[\Xi_c(2790)^{0}\to\Xi_c^{0}\gamma]/B[\Xi_c(2790)^0\to\Xi_c^{\prime +}\pi^{-}\to\Xi_c^{+}\gamma \pi^-] = 0.13 \pm 0.03 \pm 0.02$. The first quoted uncertainties are statistical and the second systematic. We find no hint of the analogous decays of the $\Xi_c(2815)^+$ and $\Xi_c(2790)^+$ baryons and set upper limits at the 90% confidence level of: $B[\Xi_c(2815)^{+}\to\Xi_c^{+}\gamma]/B[\Xi_c(2815)^+\to\Xi_c(2645)^0\pi^+\to\Xi_c^+\pi^-\pi^+] < 0.09,$ and $B[\Xi_c(2790)^{+}\to\Xi_c^{+}\gamma]/B[\Xi_c(2790)^+\to\Xi_c^{\prime 0}\pi^{+}\to\Xi_c^{0}\gamma \pi^+] < 0.06.$ Approximate values of the partial widths of the decays are extracted, which can be used to discriminate between models of the underlying quark structure of these excited states.

We report results from a study of the spin and parity of $\Xi_{c}(2970)^+$ using a $980~\mathrm{fb^{-1}}$ data sample collected by the Belle detector at the KEKB asymmetric-energy $e^{+}e^{-}$ collider. The decay angle distributions in the chain $\Xi_{c}(2970)^+ \to \Xi_c(2645)^{0}\pi^{+} \to \Xi_c^{+}\pi^{-}\pi^{+}$ are analyzed to determine the spin of this charmed-strange baryon. The angular distributions strongly favor the $\Xi_{c}(2970)^+$ spin $J =1/2$ over $3/2$ or $5/2$, under an assumption that the lowest partial wave dominates in the decay. We also measure the ratio of $\Xi_{c}(2970)^+$ decay branching fractions $R={\mathcal{B}[ \Xi_{c}(2970)^+ \to \Xi_c(2645)^{0}\pi^{+} ]} / { \mathcal{B}[ \Xi_{c}(2970)^+ \to \Xi_c^{\prime0}\pi^{+} ]} =1.67 \pm 0.29\mathrm{(stat.)}^{ +0.15}_{ -0.09}\mathrm{(syst.)} \pm 0.25\mathrm{(IS)}$, where the last uncertainty is due to possible isospin-symmetry-breaking effects. This $R$ value favors the spin-parity $J^P=1/2^+$ with the spin of the light-quark degrees of freedom $s_{l}=0$. This is the first determination of the spin and parity of a charmed-strange baryon.

Cardiovascular waveforms contain information for clinical diagnosis. By "learning" and organizing the subtle change of waveform morphology from large amounts of raw waveform data, unsupervised manifold learning helps delineate a high-dimensional structure and display it as a novel three-dimensional (3D) image. We investigate the electrocardiography (ECG) waveform for ischemic heart disease and arterial blood pressure (ABP) waveform in dynamic vasoactive episodes. We model each beat or pulse to be a point lying on a manifold, like a surface, and use the diffusion map (DMap) to establish the relationship among those pulses. For ECG datasets, first we analyzed the non-ST-elevation ECG waveform distribution from unstable angina to healthy control, and we investigated intraoperative ST-elevation ECG waveforms to show the dynamic ECG waveform changes. For ABP datasets, we analyzed waveforms collected under endotracheal intubation and administration of vasodilator. To quantify the dynamic separation, we applied the support vector machine (SVM) analysis and the trajectory analysis. For the non-ST-elevation ECG, a hierarchical tree structure comprising consecutive ECG waveforms spanning from unstable angina to healthy control is presented in the 3D image (accuracy=97.6%, macro-F1=96.1%). The DMap helps quantify and visualize the evolving direction of intraoperative ST-elevation myocardial episode in a 1-hour period (accuracy=97.58%, macro-F1=96.06%). The ABP waveform analysis of Nicardipine administration shows inter-individual difference (accuracy=95.01%, macro-F1=96.9%) and their common directions from intra-individual moving trajectories. The dynamic change of the ABP waveform during endotracheal intubation shows a loop-like trajectory structure, which can be further divided using the knowledge obtained from Nicardipine. The 3D images provide clues of underneath physiological mechanisms.