Search

This study investigates air pollution in central Taiwan, focusing on key pollutants, including SO$_2$, NO$_2$, PM$_{10}$, and PM$_{2.5}$. We use non-negative matrix factorization (NMF) to reduce data dimensionality, followed by wind direction analysis and speed to trace pollution sources. Our findings indicate that PM$_{2.5}$ and NO$_2$ levels are primarily influenced by local sources, while SO$_2$ levels are more affected by transboundary factors. For PM$_{10}$, contributions from domestic and transboundary sources are nearly equal., Comment: 25 pages, 29 figures

The dynamical quantum phase transition (DQPT) is an important concept in non-equilibrium critical phenomena, however, its relation to the equilibrium quantum phase transition (EQPT) remains obscure. Substantial evidences have suggested that quenching across the underlying equilibrium phase boundary is neither a sufficient nor a necessary condition for the existence of the DQPT. In this work, we give a necessary and sufficient condition for the occurrence of DQPTs in two-band systems by introducing the quench fidelity, which is defined as the fidelity between the ground state wavefunctions of the pre-quench and post-quench Hamiltonian, and elaborate it by taking one-dimensional anisotropic XY model as an example. The relation between EQPTs and DQPTs is analyzed in detail in terms of the quench fidelity., Comment: 9 pages, 6 figures

Synthesizing realistic videos of talking faces under custom lighting conditions and viewing angles benefits various downstream applications like video conferencing. However, most existing relighting methods are either time-consuming or unable to adjust the viewpoints. In this paper, we present the first real-time 3D-aware method for relighting in-the-wild videos of talking faces based on Neural Radiance Fields (NeRF). Given an input portrait video, our method can synthesize talking faces under both novel views and novel lighting conditions with a photo-realistic and disentangled 3D representation. Specifically, we infer an albedo tri-plane, as well as a shading tri-plane based on a desired lighting condition for each video frame with fast dual-encoders. We also leverage a temporal consistency network to ensure smooth transitions and reduce flickering artifacts. Our method runs at 32.98 fps on consumer-level hardware and achieves state-of-the-art results in terms of reconstruction quality, lighting error, lighting instability, temporal consistency and inference speed. We demonstrate the effectiveness and interactivity of our method on various portrait videos with diverse lighting and viewing conditions., Comment: Accepted to CVPR 2024 (Highlight). Project page: http://geometrylearning.com/VideoRelighting

Anderson localization and non-Hermitian skin effect are two paradigmatic wave localization phenomena, resulting from wave interference and the intrinsic non-Hermitian point gap, respectively. In this study, we unveil a novel localization phenomenon associated with long-range asymmetric coupling, termed scale-tailored localization, where the number of induced localized modes and their localization lengths scale exclusively with the coupling range. We show that the long-range coupling fundamentally reshapes the energy spectra and eigenstates by creating multiple connected paths on the lattice. Furthermore, we present experimental observations of scale-tailored localization in non-Hermitian electrical circuits utilizing adjustable voltage followers and switches. The circuit admittance spectra possess separate point-shaped and loop-shaped components in the complex energy plane, corresponding respectively to skin modes and scale-tailored localized states. Our findings not only expand and deepen the understanding of peculiar effects induced by non-Hermiticity but also offer a feasible experimental platform for exploring and controlling wave localizations., Comment: 10+16 pages, 5+11 figures

We consider the certification of temporal quantum correlations using the pseudo-density matrix (PDM), an extension of the density matrix to the time domain, where negative eigenvalues are key indicators of temporal correlations. Conventional methods for detecting these correlations rely on PDM tomography, which often involves excessive redundant information and requires exponential resources. In this work, we develop an efficient protocol for temporal correlation detection by virtually preparing the PDM within a single time slice and estimating its second-order moments using randomized measurements. Through sample complexity analysis, we demonstrate that our protocol requires only a constant number of measurement bases, making it particularly advantageous for systems utilizing ensemble average measurements, as it maintains constant runtime complexity regardless of the number of qubits. We experimentally validate our protocol on a nuclear magnetic resonance platform, a typical thermodynamic quantum system, where the experimental results closely align with theoretical predictions, confirming the effectiveness of our protocol., Comment: 19 pages, 11 figures, comments are welcome

Multi-modality image fusion aims to integrate the merits of images from different sources and render high-quality fusion images. However, existing feature extraction and fusion methods are either constrained by inherent local reduction bias and static parameters during inference (CNN) or limited by quadratic computational complexity (Transformers), and cannot effectively extract and fuse features. To solve this problem, we propose a dual-branch image fusion network called Tmamba. It consists of linear Transformer and Mamba, which has global modeling capabilities while maintaining linear complexity. Due to the difference between the Transformer and Mamba structures, the features extracted by the two branches carry channel and position information respectively. T-M interaction structure is designed between the two branches, using global learnable parameters and convolutional layers to transfer position and channel information respectively. We further propose cross-modal interaction at the attention level to obtain cross-modal attention. Experiments show that our Tmamba achieves promising results in multiple fusion tasks, including infrared-visible image fusion and medical image fusion. Code with checkpoints will be available after the peer-review process.

Quasiperiodic mosaic systems have attracted significant attention due to their unique spectral properties with exactly known mobility edges, which do not vanish even in the large quasiperiodic potential strength region, although the width of energy window of extended states becomes very narrow and decreases with the increase of strength of the quasiperiodic potential.In this work we study the dynamics of a quasiperiodic mosaic lattice and unravel its peculiar dynamical properties. By scrutinizing the expansion dynamics of wave packet and the evolution of density distribution, we unveil that the long-time density distribution display obviously different behaviors at odd and even sites in the large quasiperiodic potential strength region. Particularly, the time scale of dynamics exhibits an inverse relationship with the quasiperiodic potential strength. To understand these behaviors, we derive an effective Hamiltonian in the large quasiperiodic potential strength region, which is composed of decoupled Hamiltonians defined on the odd and even sites, respectively. While all eigenstates of the effective Hamiltonian defined on even sites are localized, the eigenstates of effective Hamiltonian defined on odd sites include both localized and extended eigenstates. Our results demonstrate that the effective Hamiltonian can describe the dynamical behaviors well in the large quasiperiodic potential strength region and provides an intuitive framework for understanding the peculiar dynamical behaviors in the quasiperiodic mosaic lattice., Comment: 11 pages, 7 figures

Manually annotating instruction data for large language models is difficult, costly, and hard to scale. Meanwhile, current automatic annotation methods typically rely on distilling synthetic data from proprietary LLMs, which not only limits the upper bound of the quality of the instruction data but also raises potential copyright issues. In this paper, we propose REInstruct, a simple and scalable method to automatically build instruction data from an unlabeled corpus without heavy reliance on proprietary LLMs and human annotation. Specifically, REInstruct first selects a subset of unlabeled texts that potentially contain well-structured helpful and insightful content and then generates instructions for these texts. To generate accurate and relevant responses for effective and robust training, REInstruct further proposes a rewriting-based approach to improve the quality of the generated instruction data. By training Llama-7b on a combination of 3k seed data and 32k synthetic data from REInstruct, fine-tuned model achieves a 65.41\% win rate on AlpacaEval leaderboard against text-davinci-003, outperforming other open-source, non-distilled instruction data construction methods. The code is publicly available at \url{https://github.com/cs32963/REInstruct}., Comment: Accepted by ACL2024 Findings

Sequential recommendation leverages interaction sequences to predict forthcoming user behaviors, crucial for crafting personalized recommendations. However, the true preferences of a user are inherently complex and high-dimensional, while the observed data is merely a simplified and low-dimensional projection of the rich preferences, which often leads to prevalent issues like data sparsity and inaccurate model training. To learn true preferences from the sparse data, most existing works endeavor to introduce some extra information or design some ingenious models. Although they have shown to be effective, extra information usually increases the cost of data collection, and complex models may result in difficulty in deployment. Innovatively, we avoid the use of extra information or alterations to the model; instead, we fill the transformation space between the observed data and the underlying preferences with randomness. Specifically, we propose a novel model-agnostic and highly generic framework for sequential recommendation called sample enrichment via temporary operations on subsequences (SETO), which temporarily and separately enriches the transformation space via sequence enhancement operations with rationality constraints in training. The transformation space not only exists in the process from input samples to preferences but also in preferences to target samples. We highlight our SETO's effectiveness and versatility over multiple representative and state-of-the-art sequential recommendation models (including six single-domain sequential models and two cross-domain sequential models) across multiple real-world datasets (including three single-domain datasets, three cross-domain datasets and a large-scale industry dataset)., Comment: 12 pages, 6 figures

The ability to detect single photons has led to the advancement of numerous research fields. Although various types of single-photon detector have been developed, because of two main factors - that is, (1) the need for operating at cryogenic temperature and (2) the incompatibility with complementary metal-oxide-semiconductor (CMOS) fabrication processes - so far, to our knowledge, only Si-based single-photon avalanche diode (SPAD) has gained mainstream success and has been used in consumer electronics. With the growing demand to shift the operation wavelength from near-infrared to short-wavelength infrared (SWIR) for better safety and performance, an alternative solution is required because Si has negligible optical absorption for wavelengths beyond 1 {\mu}m. Here we report a CMOS-compatible, high-performing germanium-silicon SPAD operated at room temperature, featuring a noise-equivalent power improvement over the previous Ge-based SPADs by 2-3.5 orders of magnitude. Key parameters such as dark count rate, single-photon detection probability at 1,310 nm, timing jitter, after-pulsing characteristic time and after-pulsing probability are, respectively, measured as 19 kHz {\mu}m^2, 12%, 188 ps, ~90 ns and <1%, with a low breakdown voltage of 10.26 V and a small excess bias of 0.75 V. Three-dimensional point-cloud images are captured with direct time-of-flight technique as proof of concept. This work paves the way towards using single-photon-sensitive SWIR sensors, imagers and photonic integrated circuits in everyday life., Comment: accepted manuscript

The skin effect, characterized by the tendency of particles to accumulate at the boundaries, has been extensively studied in non-Hermitian systems. In this work, we propose an intuitive Lindbladian composed of two chains with reversed skin localization. The skin steady state is gradually delocalized as the interchain coupling increases. In the single-body scenario, it corresponds to a shift in the scaling of the Liouvillian gap $\Delta$ from $\Delta \propto N^0$ to $\Delta \propto N^{-2}$. Notably, exact diagonalization results reveal a system-size sensitivity of the single-particle Liouvillian spectrum, inherited from the non-Hermitian effective Hamiltonian's system-size sensitivity. We predict that even an arbitrarily small coupling will induce dramatic changes in the Liouvillian spectrum and steady state in the thermodynamic limit, a phenomenon we term the critical Liouvillian skin effect. Additionally, in the many-body scenario, by employing the stochastic Schr\"odinger equation to unravel the Lindblad master equation, it is revealed that the scaling behavior of steady-state entanglement changes from the area law to the logarithmic law. This work demonstrates the delocalization of both single-body and many-body skin steady states, introducing a novel mechanism for inducing entanglement transitions beyond the quantum Zeno effect., Comment: 10 pages,7 figures

We propose a novel unsupervised cross-modal homography estimation framework based on intra-modal Self-supervised learning, Correlation, and consistent feature map Projection, namely SCPNet. The concept of intra-modal self-supervised learning is first presented to facilitate the unsupervised cross-modal homography estimation. The correlation-based homography estimation network and the consistent feature map projection are combined to form the learnable architecture of SCPNet, boosting the unsupervised learning framework. SCPNet is the first to achieve effective unsupervised homography estimation on the satellite-map image pair cross-modal dataset, GoogleMap, under [-32,+32] offset on a 128x128 image, leading the supervised approach MHN by 14.0% of mean average corner error (MACE). We further conduct extensive experiments on several cross-modal/spectral and manually-made inconsistent datasets, on which SCPNet achieves the state-of-the-art (SOTA) performance among unsupervised approaches, and owns 49.0%, 25.2%, 36.4%, and 10.7% lower MACEs than the supervised approach MHN. Source code is available at https://github.com/RM-Zhang/SCPNet., Comment: Accepted by ECCV 2024

We explicitly construct unique non-equilibrium steady state (NESS) of Lindblad master equation characterized by a Gibbs ensemble $\rho_{\text{NESS}} \propto e^{-\beta \tilde{H}}$, where the effective Hamiltonian $\tilde{H}$ consists only of $U(1)$ conserved charges of the original Hamiltonian. Specifically, when the original Hamiltonian has multiple charges, it is possible to couple them with bathes at different temperature respectively, but still leads to an equilibrium state. To access the Gibbs NESS, the jump operators need to be properly chosen to fulfill quantum detailed balance condition (qDBC). These jump operators are ladder operators for $\tilde{H}$ and jump process they generate form a vertex-weighted directed acyclic graph (wDAG). By studying the XX model and Fredkin model, we showcase how the Gibbs state emerges as the unique steady state., Comment: Updated: 10 pages, 6 figures. Added journal reference. Comments are welcome!

We propose a class of general non-Hermitian quasiperiodic lattice models with exponential hoppings and analytically determine the genuine complex mobility edges by solving its dual counterpart exactly utilizing Avila's global theory. Our analytical formula unveils that the complex mobility edges usually form a loop structure in the complex energy plane. By shifting the eigenenergy a constant $t$, the complex mobility edges of the family of models with different hopping parameter $t$ can be described by a unified formula, formally independent of $t$. By scanning the hopping parameter, we demonstrate the existence of a type of intriguing flagellate-like spectra in complex energy plane, in which the localized states and extended states are well separated by the complex mobility edges. Our result provides a firm ground for understanding the complex mobility edges in non-Hermitian quasiperiodic lattices., Comment: 8 pages, 2 figures

The intersection of physics-based vision and deep learning presents an exciting frontier for advancing computer vision technologies. By leveraging the principles of physics to inform and enhance deep learning models, we can develop more robust and accurate vision systems. Physics-based vision aims to invert the processes to recover scene properties such as shape, reflectance, light distribution, and medium properties from images. In recent years, deep learning has shown promising improvements for various vision tasks, and when combined with physics-based vision, these approaches can enhance the robustness and accuracy of vision systems. This technical report summarizes the outcomes of the Physics-Based Vision Meets Deep Learning (PBDL) 2024 challenge, held in CVPR 2024 workshop. The challenge consisted of eight tracks, focusing on Low-Light Enhancement and Detection as well as High Dynamic Range (HDR) Imaging. This report details the objectives, methodologies, and results of each track, highlighting the top-performing solutions and their innovative approaches., Comment: CVPR 2024 PBDL Challenges: https://pbdl-ws.github.io/pbdl2024/challenge/index.html

The Lindblad master equation (LME) describing the Markovian dynamics of the quantum open system can be understood as the evolution of the effective non-Hermitian Hamiltonian balanced with random quantum jumps. Here we investigate the balance-breaking dynamics by partly eliminating jumps from postselection experiments. To describe this dynamics, a non-linear Lindblad master equation (NLME) is derived from quantum trajectory method. However, the NLME shows significant advantages in analytical analysis over quantum trajectory method. Using the NLME, we classify the dynamics into two classes. In the trivial class, the process of reducing jumps is completely equivalent to weakening the coupling from the environment. In contrast, the nontrivial class presents more complex dynamics. We study a prototypical model within this class and demonstrate the existence of the postselected skin effect whose steady state is characterized by the accumulation of particles on one side. The steady-state distribution can be fitted by a scale-invariant tanh function which is different from the uniform distribution of LME. Furthermore, the NLME can give a reasonable framework for studying the interplay and competition between the non-Hermitian Hamiltonians and dissipative terms. We show this by capturing the characteristics of the trajectory-averaged entanglement entropy influenced by non-Hermitian skin effect and Zeno effect in the model of postselected skin effect.

The increasing demand for computational photography and imaging on mobile platforms has led to the widespread development and integration of advanced image sensors with novel algorithms in camera systems. However, the scarcity of high-quality data for research and the rare opportunity for in-depth exchange of views from industry and academia constrain the development of mobile intelligent photography and imaging (MIPI). Building on the achievements of the previous MIPI Workshops held at ECCV 2022 and CVPR 2023, we introduce our third MIPI challenge including three tracks focusing on novel image sensors and imaging algorithms. In this paper, we summarize and review the Nighttime Flare Removal track on MIPI 2024. In total, 170 participants were successfully registered, and 14 teams submitted results in the final testing phase. The developed solutions in this challenge achieved state-of-the-art performance on Nighttime Flare Removal. More details of this challenge and the link to the dataset can be found at https://mipi-challenge.org/MIPI2024/., Comment: MIPI@CVPR2024. Website: https://mipi-challenge.org/MIPI2024/

We unveil the existence of two-particle bound state in the continuum (BIC) in a one-dimensional interacting nonreciprocal lattice with a generalized boundary condition. By applying the Bethe-ansatz method, we can exactly solve the wavefunction and eigenvalue of the bound state in the continuum band, which enable us to precisely determine the phase diagrams of BIC. Our results demonstrate that the non-reciprocal hopping can delocalize the bound state and thus shrink the region of BIC. By analyzing the wavefunction, we identify the existence of two types of BICs with different spatial distributions and analytically derive the corresponding threshold values for the breakdown of BICs. The BIC with similar properties is also found to exist in another system with an impurity potential., Comment: 7 pages, 4 figures

Social media platforms such as Twitter, Reddit, and Sina Weibo play a crucial role in global communication but often encounter strict regulations in geopolitically sensitive regions. This situation has prompted users to ingeniously modify their way of communicating, frequently resorting to coded language in these regulated social media environments. This shift in communication is not merely a strategy to counteract regulation, but a vivid manifestation of language evolution, demonstrating how language naturally evolves under societal and technological pressures. Studying the evolution of language in regulated social media contexts is of significant importance for ensuring freedom of speech, optimizing content moderation, and advancing linguistic research. This paper proposes a multi-agent simulation framework using Large Language Models (LLMs) to explore the evolution of user language in regulated social media environments. The framework employs LLM-driven agents: supervisory agent who enforce dialogue supervision and participant agents who evolve their language strategies while engaging in conversation, simulating the evolution of communication styles under strict regulations aimed at evading social media regulation. The study evaluates the framework's effectiveness through a range of scenarios from abstract scenarios to real-world situations. Key findings indicate that LLMs are capable of simulating nuanced language dynamics and interactions in constrained settings, showing improvement in both evading supervision and information accuracy as evolution progresses. Furthermore, it was found that LLM agents adopt different strategies for different scenarios., Comment: Accepted by IEEE WCCI 2024

This paper reports on the NTIRE 2024 Quality Assessment of AI-Generated Content Challenge, which will be held in conjunction with the New Trends in Image Restoration and Enhancement Workshop (NTIRE) at CVPR 2024. This challenge is to address a major challenge in the field of image and video processing, namely, Image Quality Assessment (IQA) and Video Quality Assessment (VQA) for AI-Generated Content (AIGC). The challenge is divided into the image track and the video track. The image track uses the AIGIQA-20K, which contains 20,000 AI-Generated Images (AIGIs) generated by 15 popular generative models. The image track has a total of 318 registered participants. A total of 1,646 submissions are received in the development phase, and 221 submissions are received in the test phase. Finally, 16 participating teams submitted their models and fact sheets. The video track uses the T2VQA-DB, which contains 10,000 AI-Generated Videos (AIGVs) generated by 9 popular Text-to-Video (T2V) models. A total of 196 participants have registered in the video track. A total of 991 submissions are received in the development phase, and 185 submissions are received in the test phase. Finally, 12 participating teams submitted their models and fact sheets. Some methods have achieved better results than baseline methods, and the winning methods in both tracks have demonstrated superior prediction performance on AIGC.

We propose a family of exactly solvable quasiperiodic lattice models with analytical complex mobility edges, which can incorporate mosaic modulations as a straightforward generalization. By sweeping a potential tuning parameter $\delta$, we demonstrate a kind of interesting butterfly-like spectra in complex energy plane, which depicts energy-dependent extended-localized transitions sharing a common exact non-Hermitian mobility edge. Applying Avila's global theory, we are able to analytically calculate the Lyapunov exponents and determine the mobility edges exactly. For the minimal model without mosaic modulation, a compactly analytic formula for the complex mobility edges is obtained, which, together with analytical estimation of the range of complex energy spectrum, gives the true mobility edge. The non-Hermitian mobility edge in complex energy plane is further verified by numerical calculations of fractal dimension and spatial distribution of wave functions. Tuning parameters of non-Hermitian potentials, we also investigate the variations of the non-Hermitian mobility edges and the corresponding butterfly spectra, which exhibit richness of spectrum structures., Comment: 12 pages, 11 figures

Cross-spectral image guided denoising has shown its great potential in recovering clean images with rich details, such as using the near-infrared image to guide the denoising process of the visible one. To obtain such image pairs, a feasible and economical way is to employ a stereo system, which is widely used on mobile devices. Current works attempt to generate an aligned guidance image to handle the disparity between two images. However, due to occlusion, spectral differences and noise degradation, the aligned guidance image generally exists ghosting and artifacts, leading to an unsatisfactory denoised result. To address this issue, we propose a one-stage transformer-based architecture, named SGDFormer, for cross-spectral Stereo image Guided Denoising. The architecture integrates the correspondence modeling and feature fusion of stereo images into a unified network. Our transformer block contains a noise-robust cross-attention (NRCA) module and a spatially variant feature fusion (SVFF) module. The NRCA module captures the long-range correspondence of two images in a coarse-to-fine manner to alleviate the interference of noise. The SVFF module further enhances salient structures and suppresses harmful artifacts through dynamically selecting useful information. Thanks to the above design, our SGDFormer can restore artifact-free images with fine structures, and achieves state-of-the-art performance on various datasets. Additionally, our SGDFormer can be extended to handle other unaligned cross-model guided restoration tasks such as guided depth super-resolution.

We study the spread complexity in two-mode Bose-Einstein condensations and unveil that the long-time average of the spread complexity $\overline{C}_{K}$ can probe the dynamical transition from self-trapping to Josephson oscillation. When the parameter $\omega$ increases over a critical value $\omega_{c}$, we reveal that the spread complexity exhibits a sharp transition from lower to higher value, with the corresponding phase space trajectory changing from self-trapping to Josephson oscillation. Moreover, we scrutinize the eigen-spectrum and uncover the relation between the dynamical transition and the excited state quantum phase transition, which is characterized by the emergence of singularity in the density of states at critical energy $E_{c}$. In the thermodynamical limit, the cross point of $E_{c}(\omega)$ and the initial energy $E_{0}(\omega)$ determines the dynamical transition point $\omega_{c}$. Furthermore, we show that the different dynamical behavior for the initial state at a fixed point can be distinguished by the long-time average of the spread complexity, when the fixed point changes from unstable to stable. Finally, we also examine the sensitivity of $\overline{C}_{K}$ for the triple-well bosonic model which exibits the transition from chaotic dynamics to regular dynamics., Comment: 12 pages, 13 figures

When doped into a certain range of charge carrier concentrations, MoS2 departs from its pristine semiconducting character to become a strongly correlated material characterized by exotic phenomena such as charge density waves or superconductivity. However, the required doping levels are typically achieved using ionic-liquid gating or air-sensitive alkali-ion intercalation, which are not compatible with standard device fabrication processes. Here, we report on the emergence of superconductivity and a charge density wave phase in air-stable organic cation intercalated MoS2 crystals. By selecting two different molecular guests, we show that these correlated electronic phases depend dramatically on the intercalated cation, demonstrating the potential of organic ion intercalation to finely tune the properties of 2D materials. Moreover, we find that a fully developed zero-resistance state is not reached in few-nm-thick flakes, indicating the presence of three-dimensional superconductive paths which are severed by the mechanical exfoliation. We ascribe this behavior to an inhomogeneous charge carrier distribution, which we probe at the nanoscale using scanning near-field optical microscopy. Our results establish organic-ion intercalated MoS2 as a platform to study the emergence and modulation of correlated electronic phases.

This study explores the production of open-charm pentaquark molecular states, specifically $N\bar{D}^*$ and $\bar{N}\bar{D}^*$, within the $B^0 \rightarrow \bar{D}^0 p \bar{p}$ decay process. We analyze the invariant mass spectrum of $p\bar{D}^0$ and $\bar{p}\bar{D}^0$, incorporating the rescattering process calculated using a quasipotential Bethe-Salpeter equation approach. Our findings suggest the potential identification of the isoscalar $\bar{N}\bar{D}^*$ molecule with $3/2^+$, serving as the antiparticle partner of the $\Lambda_c(2940)$, in the $\bar{p}\bar{D}^0$ mass distribution. Additionally, distinctive signals of the isovector $N\bar{D}^*$ molecule with $1/2^-$ may emerge in the $p\bar{D}^0$ invariant mass distribution. We highlight the significance of the three-body decay of the bottom meson as a valuable avenue for studying open-charm molecules and advocate for increased attention and more precise experimental measurements of the $B^0 \rightarrow \bar{D}^0 p \bar{p}$ process., Comment: 8 pages, 3 figures

We present the first application of the weighted skew-spectra to analyze non-Gaussian information in galaxy survey data. Using the tree-level galaxy skew-spectra together with the one-loop power spectrum multipoles, we analyze the Sloan Digital Sky Survey (SDSS)-III Baryon Oscillation Spectroscopic Survey (BOSS) galaxy clustering data, and target our search towards the equilateral bispectrum shape of primordial non-Gaussianity. We use the Effective Field Theory model for the galaxy power spectrum and bispectrum, and account for systematic effects, such as the survey geometry. From our likelihood analysis, we find $f_{\rm NL}^{\rm equil}=-34^{+296}_{-334}$ at $68\%$ CL, consistent with previous works, while systematic errors from our treatment of the survey geometry lead to an unreliable estimation of $f_{\rm NL}^{\rm ortho}$. We further constrain the bias and counterterm parameters, while keeping the cosmology fixed to $\textit{Planck }2018$ values. As a check, we also validate our analysis pipeline using the ${\tt Nseries}$ simulation suite., Comment: 31 pages, 1 figure, 3 tables

Cross-domain sequential recommendation (CDSR) shifts the modeling of user preferences from flat to stereoscopic by integrating and learning interaction information from multiple domains at different granularities (ranging from inter-sequence to intra-sequence and from single-domain to cross-domain). In this survey, we first define the CDSR problem using a four-dimensional tensor and then analyze its multi-type input representations under multidirectional dimensionality reductions. Following that, we provide a systematic overview from both macro and micro views. From a macro view, we abstract the multi-level fusion structures of various models across domains and discuss their bridges for fusion. From a micro view, focusing on the existing models, we first discuss the basic technologies and then explain the auxiliary learning technologies. Finally, we exhibit the available public datasets and the representative experimental results as well as provide some insights into future directions for research in CDSR., Comment: Accepted to the IJCAI 2024 Survey Track

Respiratory diseases represent one of the most significant economic burdens on healthcare systems worldwide. The variation in the increasing number of cases depends greatly on climatic seasonal effects, socioeconomic factors, and pollution. Therefore, understanding these variations and obtaining precise forecasts allows health authorities to make correct decisions regarding the allocation of limited economic and human resources. This study aims to model and forecast weekly hospitalizations due to respiratory conditions in seven regional hospitals in Costa Rica using four statistical learning techniques (Random Forest, XGboost, Facebook's Prophet forecasting model, and an ensemble method combining the above methods), along with 22 climate change indices and aerosol optical depth as an indicator of pollution. Models are trained using data from 2000 to 2018 and are evaluated using data from 2019 as testing data. Reliable predictions are obtained for each of the seven regional hospitals

Systems of interacting bosons in double-well potentials, modeled by two-site Bose-Hubbard models, are of significant theoretical and experimental interest and attracted intensive studies in contexts ranging from many-body physics and quantum dynamics to the onset of quantum chaos. In this work we systematically study a kicked two-site Bose-Hubbard model (Bose-Hubbard dimer) with the on-site potential difference being periodically modulated. Our model can be equivalently represented as a kicked Lipkin-Meshkov-Glick model and thus displays different dynamical behaviors from the kicked top model. By analyzing spectral statistics of Floquet operator, we unveil that the system undergoes a transition from regularity to chaos with increasing the interaction strength. Then based on semiclassical approximation and the analysis of R\'{e}nyi entropy of coherent states in the basis of Floquet operator eigenstates, we reveal the local chaotic features of our model, which indicate the existence of integrable islands even in the deep chaotic regime. The semiclassical analysis also suggests that the system in chaotic regime may display different dynamical behavior depending on the choice of initial states. Finally, we demonstrate that dynamical signatures of chaos can be manifested by studying dynamical evolution of local operators and out of time order correlation function as well as the entanglement entropy. Our numerical results exhibit the richness of dynamics of the kicked Bose-Hubbard dimer in both regular and chaotic regimes as the initial states are chosen as coherent spin states located in different locations of phase space., Comment: 13 pages, 14 figures