Deep Learning-Based Stock Price Prediction Using LSTM Model

The stock market is a vital component of the broader financial system,with its dynamics closely linked to economic growth.The challenges associated with analyzing and forecasting stock prices have persisted since the inception of financial markets.By examining historical transaction data,latent opportunities for profit can be uncovered,providing valuable insights for both institutional and individual investors to make more informed decisions.This study focuses on analyzing historical transaction data from four banks to predict closing price trends.Various models,including decision trees,random forests,and Long Short-Term Memory(LSTM)networks,are employed to forecast stock price movements.Historical stock transaction data serves as the input for training these models,which are then used to predict upward or downward stock price trends.The study’s empirical results indicate that these methods are effective to a degree in predicting stock price movements.The LSTM-based deep neural network model,in particular,demonstrates a commendable level of predictive accuracy.This conclusion is reached following a thorough evaluation of model performance,highlighting the potential of LSTM models in stock market forecasting.The findings offer significant implications for advancing financial forecasting approaches,thereby improving the decision-making capabilities of investors and financial institutions.

Battery Separators Functionalized with Edge-Rich MoS2/C Hollow Microspheres for the Uniform Deposition of Li2S in High-Performance Lithium-Sulfur Batteries

As promising energy storage systems,lithium-sulfur(Li-S)batteries have attracted significant attention because of their ultra-high energy densities.However,Li-S battery suffers problems related to the complex phase conversion that occurs during the charge-discharge process,particularly the deposition of solid Li2S from the liquid-phase polysulfides,which greatly limits its practical application.In this paper,edge-rich MoS2/C hollow microspheres(Edg-MoS2/C HMs)were designed and used to functionalize separator for Li-S battery,resulting in the uniform deposition of Li2S.The microspheres were fabricated through the facile hydrothermal treatment of MoO3-aniline nanowires and a subsequent carbonization process.The obtained Edg-MoS2/C HMs have a strong chemical absorption capability and high density of Li2S binding sites,and exhibit excellent electrocatalytic performance and can effectively hinder the polysulfide shuttle effect and guide the uniform nucleation and growth of Li2S.Furthermore,we demonstrate that the Edg-MoS2/C HMs can effectively regulate the deposition of Li2S and significantly improve the reversibility of the phase conversion of the active sulfur species,especially at high sulfur loadings and high C-rates.As a result,a cell containing a separator functionalized with Edg-MoS2/C HMs exhibited an initial discharge capacity of 935 mAh g-1 at 1.0 C and maintained a capacity of 494 mAh g-1 after 1000 cycles with a sulfur loading of 1.7 mg cm-2.Impressively,at a high sulfur loading of 6.1 mg cm-2 and high rate of 0.5 C,the cell still delivered a high reversible discharge capacity of 478 mAh g-1 after 300 cycles.This work provides fresh insights into energy storage systems related to complex phase conversions.

Demand analysis of an intelligent medication administration system for older adults with chronic diseases based on the Kano model

Objectives Gerontechnology has great potential in promoting older adults’well-being.With the accelerated aging process,gerontechnology has a promising market prospect.However,most technological developers and healthcare professionals attached importance to products’effectiveness,and ignored older adults’demands and user experience,which reduced older adults'adoption intention of gerontechnology use.The inclusion of older adults in the design process of technologies is essential to maximize the effect.This study explored older adults’demands for a self-developed intelligent medication administration system and proposed optimization schemes,thus providing reference to developing geriatric-friendly technologies and products.Methods A cross-sectional survey was conducted to explore older adults’technological demands for the self-developed intelligent medication administration system,and data were analyzed based on the Kano model.A self-made questionnaire was administered from July 2020 to October 2020 after participants used this system for two weeks.The study was registered with the Chinese Clinical Trial Registry(ChiCTR2000040644).Results A total of 354 older adults participated in the survey.Four items,namely larger font size,simpler operation process,scheduled medication reminders and reliable hardware,were classified as must-be attributes;three items,namely searching drug instructions through WeChat,more sensitive system and longer battery life,as attractive attributes;one item,viewing disease-related information through WeChat,as the one-dimensional attribute;and the rest were indifferent attributes,including simple and beautiful displays,blocking advertisements automatically,providing user privacy protection protocol,viewing personal medical information only by logged-in users,recording all the medications,ordering medications through WeChat.The satisfaction values were between 0.24 and 0.69,and dissatisfaction values were between 0.06 and 0.94.Conclusion This study suggested that older adults had personalized technology demands.Including their technological demands and desire may assist in decreasing the digital divide and promoting the satisfaction of e-health and/or m-health.Based on older adults’demands,our study proposed optimization schemes of the intelligent medication administration system,which may help developers design geriatric-friendly intelligent products and nurses to perform older adults-centered and efficient medication management.

Decorating ketjen black with ultra-small Mo_(2)C nanoparticles to enhance polysulfides chemisorption and redox kinetics for lithium-sulfur batteries

The low sulfur utilization and fast capacity fading resulting from the sluggish redox reaction and abominable polysulfides shuttle greatly hinder the practical applications of lithium-sulfur(Li-S) batteries.Herein, we develop a facile "in-situ growth" method to decorate ultra-small Mo2 C nanoparticles(USMo2 C) on the surface of Ketjen Black(KB) to functionalize the commercial polypropylene(PP) separators,which can accelerate the redox kinetics of lithium polysulfides conversion and effectively increase the utilization of sulfur for Li-S batteries. Importantly, the US-Mo2 C nanoparticles have abundant sites for chemical adsorption towards polysulfides and the conductive carbon networks of KB have cross-linked pore channels, which can promote electron transport and provide physical barrier and volume expansion space for polysulfides. Due to the combined effects of the US-Mo2 C and KB, Li-S cells employing the multifunctional PP separators modified with KB/US-Mo2 C composite(KB/US-Mo2 C@PP) exhibit a high specific capacity(1212.8 mAh g^(-1) at 0.2 C), and maintain a reversible capacity of 1053.3 m Ah g^(-1) after 100 cycles.More importantly, the KB/US-Mo2 C@PP cells with higher sulfur mass loading of 4.9 mg cm^(-2) have superb areal capacity of 2.3 mAh cm^(-2). This work offers a novel and promising perspective for high-performance Li-S batteries from both the shuttle effect and the complex polysulfides conversion.

基于免训练自监督式神经网络算法实现压缩超快成像高保真图像重构

压缩超快成像(compressed ultrafast photography,CUP)是目前最快的被动式单次超快光学成像技术,它通过数据获取和图像重构两个步骤实现超快事件的捕捉,已发展为记录不可逆或难以重复超快事件的一种有力工具,且能够探测荧光动力学等自发光瞬态场景.然而,传统的迭代优化型算法在图像重构上的保真度较低,而端到端型深度学习算法则严重依赖训练数据,训练复杂度高、通用性不足,这限制了CUP对超快现象进行高空间分辨率的观测.为此,我们开发了一种新型的免训练自监督式神经网络算法,其通过即插即用框架(plug-and-play,PnP)与深度图像先验(deep image prior,DIP)的结合可实现CUP的低复杂度高保真图像重建,简称为PnP-DIP算法.PnP-DIP基于交替方向乘子法(alternating direction method of multipliers,ADMM),利用DIP和PnP去噪器解决图像恢复子问题,可以在防止数据过拟合和噪声累积的同时,显著提高图像重建的精度与收敛速度.通过数值模拟,我们理论上证明了PnP-DIP算法在重构原始动态信息方面相比传统ADMM算法具有更高的保真度.同时,我们分别利用PnPDIP对自主研制CUP系统观测的皮秒激光脉冲和X射线闪烁体的时空强度演化数据进行重构,实验上验证了其优越的图像重构性能.这一研究有望推动CUP在高时空分辨观测需求中的应用,并为超快动力学的实时探测作出重大贡献.

Recent advances in long-persistent luminescence materials based on host–guest architecture

Organic long-persistent luminescence(LPL)materials,featuring low preparation cost,eco-friendly synthesis,and easy modification of functional groups,have exhibited extensive applications in information encryption,anti-counterfeiting,and biological imaging.Several design strategies including crystallization-inducement,H-aggregation,and host-guest doping to enhance persistent-room-temperature phosphorescence(RTP)effect by precisely controlling intersystem crossing(ISC)constant and suppressing nonradiative decay rates,those are important strategies to enable LPL performance.Among the strategies,researchers have made several efforts to enhance persistent-RTP effect by host-guest interaction,in which the host matrices provide a rigid environment for phosphor guest molecules.The interaction of the luminescent guest molecules with the host matrix can effectively reduce the vibration and rotation of the luminescent molecules,and suppress the non-radiative inactivation,thereby improving the phosphorescence quantum yield.This review aims to summarize several design strategies of pure organic LPL materials based on persistent-RTP effect through host-guest interaction,and describe some applications of pure organic LPL materials in different fields.

Regulated extravascular microenvironment via reversible thermosensitive hydrogel for inhibiting calcium influx and vasospasm

Arterial vasospasm after microsurgery can cause severe obstruction of blood flow manifested as low tissue temperature,leading to tissue necrosis.The timely discovery and synchronized treatment become pivotal.In this study,a reversible,intelligent,responsive thermosensitive hydrogel system is constructed employing both the gel–sol transition and the sol–gel transition.The“reversible thermosensitive(RTS)”hydrogel loaded with verapamil hydrochloride is designed to dynamically and continuously regulate the extravascular microenvi-ronment by inhibiting extracellular calcium influx.After accurate implantation and following in situ gelation,the RTS hydrogel reverses to the sol state causing massive drug release to inhibit vasospasm when the tissue tem-perature drops to the predetermined transition temperature.Subsequent restoration of the blood supply allevi-ates further tissue injury.Before the temperature drops,the RTS hydrogel maintains the gel state as a sustained-release reservoir to prevent vasospasm.The inhibition of calcium influx and vasospasm in vitro and in vivo is demonstrated using vascular smooth muscle cells,mice mesenteric arterial rings,and vascular ultrasonic Doppler detection.Subsequent animal experiments demonstrate that RTS hydrogel can promote tissue survival and alleviate tissue injury responding to temperature change.Therefore,this RTS hydrogel holds therapeutic po-tential for diseases requiring timely detection of temperature change.

A Ni/Ni_(2)P heterostructure in modified porous carbon separator for boosting polysulfide catalytic conversion

The intrinsic sluggish conversion kinetics and severe shuttle effect in lithium-sulfur(Li-S)batteries are responsible for their poor reversible capacity and cycling longevity,which have greatly hindered their practical applications.To address these drawbacks,herein,we design and construct a heterostructured Ni/Ni_(2)P embedded in a mesoporous carbon nanosphere composite(Ni/Ni_(2)P-MCN)for boosting polysulfide catalytic conversion in Li-S batteries.The Ni/Ni_(2)PMCN-modified separator could not only prevent the shuttle effect significantly through abundant chemical adsorptive sites,but also demonstrate superior catalytic reactivities for the conversion of polysulfides.More importantly,the conductive carbon matrix with an exposed mesoporous structure can serve as an effective physical barrier to accommodate deposited insoluble Li_(2)S.Consequently,the cells with the Ni/Ni_(2)P-MCN-modified separator exhibit greatly boosted rate capability(431 mA h g^(-1) at 5 C)and cycling stability(a capacity decay of 0.031% per cycle after 1500 cycles).Even at an enhanced sulfur loading of 4.2 mg cm^(-2),a stable and superior areal capacity(about 3.5 mA h cm^(-2))has been demonstrated.We envision that the unique Ni/Ni_(2)P heterostructure in the porous carbon matrix could offer great potential for highperformance and sustained energy storage devices.

利用"Find-eat"策略通过受体功能化凋亡小体纳米囊泡靶向内皮细胞

慢性创面的缺氧微环境导致血管化不足,进而使得伤口延迟愈合.内皮细胞作为组成新生血管的关键细胞,在血管化过程中发挥着核心作用.本研究构建了CX3CR1功能化凋亡小体纳米囊泡(nABs).“Find-eat”策略通过受体-配体组合,靶向低氧微环境中高表达CX3CR1的内皮细胞(ECs),从而放大“Find-eat”信号,促进血管生成.作者首先通过化学诱导脂肪干细胞(ADSCs)凋亡获得ABs,然后通过优化的低渗处理-温和超声-药物混合-挤压一系列处理步骤,获得载有去铁胺(DFO)的功能化凋亡小体纳米囊泡.体外实验显示ABs纳米囊泡生物相容性良好;可以通过CX3CL1/CX3CR1诱导缺氧微环境中的内皮细胞实现对其的“Find-eat”,进而促进内皮细胞的增殖、迁移、成管能力.体内实验中,ABs纳米囊泡可以促进伤口快速闭合,并可通过释放“Find-eat”信号靶向内皮细胞的同时实现促血管药物的缓释,促进糖尿病大鼠创面的新生血管形成.这种能够通过释放双重信号靶向内皮细胞并实现促血管药物缓释的ABs纳米囊泡为慢性糖尿病创面的愈合提供了一种新思路.