Search

Finance is a particularly difficult playground for deep reinforcement learning. However, establishing high-quality market environments and benchmarks for financial reinforcement learning is challenging due to three major factors, namely, low signal-to-noise ratio of financial data, survivorship bias of historical data, and model overfitting in the backtesting stage. In this paper, we present an openly accessible FinRL-Meta library that has been actively maintained by the AI4Finance community. First, following a DataOps paradigm, we will provide hundreds of market environments through an automatic pipeline that collects dynamic datasets from real-world markets and processes them into gym-style market environments. Second, we reproduce popular papers as stepping stones for users to design new trading strategies. We also deploy the library on cloud platforms so that users can visualize their own results and assess the relative performance via community-wise competitions. Third, FinRL-Meta provides tens of Jupyter/Python demos organized into a curriculum and a documentation website to serve the rapidly growing community. FinRL-Meta is available at: https://github.com/AI4Finance-Foundation/FinRL-Meta, Comment: NeurIPS 2022 Datasets and Benchmarks. 36th Conference on Neural Information Processing Systems Datasets and Benchmarks Track

Previous investigations of 4$H_b$-TaS$_{2-2x}$Se$_{2x}$ mainly focused on the direct competition between superconductivity and charge density wave (CDW). However, the superconductivity itself, although has been prominently enhanced by isovalent Se substitution, has not been adequately investigated. Here, we performed a detailed electrical transport measurement down to 0.1 K on a series of 4$H_b$-TaS$_{2-2x}$Se$_{2x}$ single crystals. A systematic fitting of the temperature-dependent resistance demonstrates that the decreased Debye temperatures ($\Theta_{D}$) and higher electron-phonon coupling constants ($\lambda_{e-p}$) at the optimal Se doping content raise the superconducting transition temperature ($T_c$). Additionally, we discovered that the incorporation of Se diminishes the degree of anisotropy of the superconductivity in the highly layered structure. More prominently, a comprehensive analysis of the vortex liquid phase region reveals that the optimally doped sample deviates from the canonical 2D Tinkham prediction but favors a linear trend with the variation of the external magnetic field. These findings emphasize the importance of interlayer interaction in this segregated superconducting-Mott-insulating system., Comment: 11 pages, 5 figures

Most existing pre-trained language representation models (PLMs) are sub-optimal in sentiment analysis tasks, as they capture the sentiment information from word-level while under-considering sentence-level information. In this paper, we propose SentiWSP, a novel Sentiment-aware pre-trained language model with combined Word-level and Sentence-level Pre-training tasks. The word level pre-training task detects replaced sentiment words, via a generator-discriminator framework, to enhance the PLM's knowledge about sentiment words. The sentence level pre-training task further strengthens the discriminator via a contrastive learning framework, with similar sentences as negative samples, to encode sentiments in a sentence. Extensive experimental results show that SentiWSP achieves new state-of-the-art performance on various sentence-level and aspect-level sentiment classification benchmarks. We have made our code and model publicly available at https://github.com/XMUDM/SentiWSP., Comment: Accepted to EMNLP 2022

In this work, we study the simple yet universally applicable case of reward shaping in value-based Deep Reinforcement Learning (DRL). We show that reward shifting in the form of the linear transformation is equivalent to changing the initialization of the $Q$-function in function approximation. Based on such an equivalence, we bring the key insight that a positive reward shifting leads to conservative exploitation, while a negative reward shifting leads to curiosity-driven exploration. Accordingly, conservative exploitation improves offline RL value estimation, and optimistic value estimation improves exploration for online RL. We validate our insight on a range of RL tasks and show its improvement over baselines: (1) In offline RL, the conservative exploitation leads to improved performance based on off-the-shelf algorithms; (2) In online continuous control, multiple value functions with different shifting constants can be used to tackle the exploration-exploitation dilemma for better sample efficiency; (3) In discrete control tasks, a negative reward shifting yields an improvement over the curiosity-based exploration method.

Here, we have grown single crystals of LaTe2-xSbx (x=0 and 0.4) with continuously adjustable CDW. High-pressure x-ray diffraction show LaTe2 does not undergo phase transition and keep robust below 40 GPa. In-situ high-pressure electrical measurements show LaTe2-xSbx undergo semiconductor-metal-superconductivity transition at 4.6 and 2.5 GPa, respectively. With the doping of Sb, the highest Tc increases from 4.6 to 6.5 K. Theoretical calculations reveal that the CDW has been completely suppressed and the calculated Tc is about 2.97 K at 4.5 GPa, consistent with the measured value. Then, the pressure-induced superconductivity in LaTe2-xSbx can be explained in the framework of the BCS theory., Comment: 11 pages, 5 figures

Due to the instability and limitations of unimodal biometric systems, multimodal systems have attracted more and more attention from researchers. However, how to exploit the independent and complementary information between different modalities remains a key and challenging problem. In this paper, we propose a multimodal biometric fusion recognition algorithm based on fingerprints and finger veins (Fingerprint Finger Veins-Channel Spatial Attention Fusion Module, FPV-CSAFM). Specifically, for each pair of fingerprint and finger vein images, we first propose a simple and effective Convolutional Neural Network (CNN) to extract features. Then, we build a multimodal feature fusion module (Channel Spatial Attention Fusion Module, CSAFM) to fully fuse the complementary information between fingerprints and finger veins. Different from existing fusion strategies, our fusion method can dynamically adjust the fusion weights according to the importance of different modalities in channel and spatial dimensions, so as to better combine the information between different modalities and improve the overall recognition performance. To evaluate the performance of our method, we conduct a series of experiments on multiple public datasets. Experimental results show that the proposed FPV-CSAFM achieves excellent recognition performance on three multimodal datasets based on fingerprints and finger veins.

MPX3 (M=metals, X=S or Se) represents a large family of van der Waals (vdW) materials featuring with P-P dimers of ~2.3 {\AA} separation. Its electrical transport property and structure can hardly be tuned by the intentional chemical doping and ionic intercalation. Here, we employ an entropy-enhancement strategy to successfully obtain a series of medium-entropy compounds MEPSe3 (ME=Fe, Mn, Cd and In), in which the electrical and magnetic properties changed simultaneously. Lone-pair electrons of P emerge due to the dissociation of the dimers as evidenced by a 35% elongation in the P-P interatomic distance. The band gap widens from 0.1 eV to 0.7 eV by this dissociation. Under external physical pressure up to ~50 GPa, a giant collapse of up to 15% in the c-axis happens, which is in contrast to the in-plane shrinkage of their counterparts Fe/MnPSe3. It leads to the recombination of P3- with lone pair electrons into a P-P dimer and the smallest bulk modulus of 28 GPa in MPX3. The MEPSe3 transits from a spin-glass insulator to metal, and to superconductor, which is rarely observed in the MPX3. Our findings highlight the P-P dimer as an indicator to probe diverse electronic structure and the effectiveness of entropy-enhancement in materials science., Comment: 10 pages, 4 figures

In this work, we propose using the mixture density network (MDN) to estimate cosmological parameters. We test the MDN method by constraining parameters of the $\Lambda$CDM and $w$CDM models using Type Ia supernovae and the power spectra of the cosmic microwave background. We find that the MDN method can achieve the same level of accuracy as the Markov Chain Monte Carlo method, with a slight difference of $\mathcal{O}(10^{-2}\sigma)$. Furthermore, the MDN method can provide accurate parameter estimates with $\mathcal{O}(10^3)$ forward simulation samples, which are useful for complex and resource-consuming cosmological models. This method can process either one data set or multiple data sets to achieve joint constraints on parameters, extendable for any parameter estimation of complicated models in a wider scientific field. Thus, the MDN provides an alternative way for likelihood-free inference of parameters., Comment: 17 pages, 4 tables, 15 figures, match the published version. The code repository is available at https://github.com/Guo-Jian-Wang/mdncoper

The cosmic microwave background (CMB), carrying the inhomogeneous information of the very early universe, is of great significance for understanding the origin and evolution of our universe. However, observational CMB maps contain serious foreground contaminations from several sources, such as galactic synchrotron and thermal dust emissions. Here, we build a deep convolutional neural network (CNN) to recover the tiny CMB signal from various huge foreground contaminations. Focusing on the CMB temperature fluctuations, we find that the CNN model can successfully recover the CMB temperature maps with high accuracy, and that the deviation of the recovered power spectrum $C_\ell$ is smaller than the cosmic variance at $\ell>10$. We then apply this method to the current Planck observation, and find that the recovered CMB is quite consistent with that disclosed by the Planck collaboration, which indicates that the CNN method can provide a promising approach to the component separation of CMB observations. Furthermore, we test the CNN method with simulated CMB polarization maps based on the CMB-S4 experiment. The result shows that both the EE and BB power spectra can be recovered with high accuracy. Therefore, this method will be helpful for the detection of primordial gravitational waves in current and future CMB experiments. The CNN is designed to analyze two-dimensional images, thus this method is not only able to process full-sky maps, but also partial-sky maps. Therefore, it can also be used for other similar experiments, such as radio surveys like the Square Kilometer Array., Comment: 20 pages, 25 figures, and 3 tables, updated citations in section 1. The code repository is available at https://github.com/Guo-Jian-Wang/cmbNNCS

This Snowmass 2021 White Paper describes the Cosmic Microwave Background Stage 4 project CMB-S4, which is designed to cross critical thresholds in our understanding of the origin and evolution of the Universe, from the highest energies at the dawn of time through the growth of structure to the present day. We provide an overview of the science case, the technical design, and project plan., Comment: Contribution to Snowmass 2021. arXiv admin note: substantial text overlap with arXiv:1908.01062, arXiv:1907.04473

Two-dimensional electron gas (2DEG) formed at the interface between two insulating oxides offers an opportunity for fundamental research and device applications. Binary alkaline earth metal oxides possess compatible lattice constants with both silicon and perovskite oxides, exhibiting an enormous potential to bridging those two materials classes for multifunctionalities. Here we report the formation of 2DEG at the interface between the rock-salt BaO and SrTiO3. The highest electron mobility reaches 69000 cm^2 V.S^-1 at 2 K, leading to the typical Shubniko de Haas (SdH) oscillations under the high magnetic fields. The presence of SdH oscillations at different field-angles reveals a quasi-two-dimensional character of the Fermi surface. The first-principles calculations suggest that the effective charge transfer from the BaO to Ti 3dxy orbital at the interfaces is responsible to the observed high carrier mobility. Our results demonstrate that the BaO/STO heterointerface is a platform for exploring the correlated quantum phases, opening a door to the low-power and mesoscopic electronic devices., Comment: 18 pages; 5 figures

This paper presents a comprehensive survey of vision-language (VL) intelligence from the perspective of time. This survey is inspired by the remarkable progress in both computer vision and natural language processing, and recent trends shifting from single modality processing to multiple modality comprehension. We summarize the development in this field into three time periods, namely task-specific methods, vision-language pre-training (VLP) methods, and larger models empowered by large-scale weakly-labeled data. We first take some common VL tasks as examples to introduce the development of task-specific methods. Then we focus on VLP methods and comprehensively review key components of the model structures and training methods. After that, we show how recent work utilizes large-scale raw image-text data to learn language-aligned visual representations that generalize better on zero or few shot learning tasks. Finally, we discuss some potential future trends towards modality cooperation, unified representation, and knowledge incorporation. We believe that this review will be of help for researchers and practitioners of AI and ML, especially those interested in computer vision and natural language processing.

We present in this paper a novel denoising training method to speedup DETR (DEtection TRansformer) training and offer a deepened understanding of the slow convergence issue of DETR-like methods. We show that the slow convergence results from the instability of bipartite graph matching which causes inconsistent optimization goals in early training stages. To address this issue, except for the Hungarian loss, our method additionally feeds ground-truth bounding boxes with noises into Transformer decoder and trains the model to reconstruct the original boxes, which effectively reduces the bipartite graph matching difficulty and leads to a faster convergence. Our method is universal and can be easily plugged into any DETR-like methods by adding dozens of lines of code to achieve a remarkable improvement. As a result, our DN-DETR results in a remarkable improvement ($+1.9$AP) under the same setting and achieves the best result (AP $43.4$ and $48.6$ with $12$ and $50$ epochs of training respectively) among DETR-like methods with ResNet-$50$ backbone. Compared with the baseline under the same setting, DN-DETR achieves comparable performance with $50\%$ training epochs. Code is available at \url{https://github.com/FengLi-ust/DN-DETR}., Comment: Extended version from CVPR 2022