LSTM在低能见度预报中的应用

低能见度天气是制约交通出行的主要因素之一，提前做好低能见度的预报工作能利于提前决策、规避风险、降低损失。由于低能见度天气是时间序列问题，针对这种问题，近年来深度学习的长短时记忆网络模型能够显著提高预报精度，因此利用多变量数据对茅台机场低能见度进行预测。首先进行气象因素分析，利用皮尔逊相关系数法来挑选相关性高的气象因子，从而减小数据量得到更利于训练的数据集。然后利用LSTM网络对多元时间序列进行建模，实现机场低能见度的预测。经过试验，模型对茅台机场能见度的准确预报率为85.43%，为机场低能见度预报提供了新的方法。     

Long memory in the high frequency cryptocurrency markets using fractal connectivity analysis: The impact of COVID-19

In this paper, we study the long memory behavior of the hourly cryptocurrency returns during the COVID-19 pandemic period. Initially, we apply different tests against the spurious long memory, with the results indicating the presence of true long memory for most cryptocurrencies. Yet, using the multivariate test, the series are found to be contaminated by level shifts or smooth trends. Then, we adopt the wavelet-based multivariate long memory approach suggested by Achard and Gannaz (2016) to model their long memory connectivity. The findings indicate a change in persistence for all series during the sample period. The fractal connectivity clustering indicates a similarity among Ethereum (ETH) and Litecoin (LTC), Monero (XMR), Bitcoin (BTC), and EOC token (EOS), while Stellar (XLM) is clustered away from the remaining series, indicating the absence of any interdependence with other crypto returns. Overall, shocks arising from COVID-19 crisis have led to changes in long-run correlation structure.     

政府审计能提升国有企业并购重组业绩承诺可靠性吗？——基于审计署央企审计结果公告的经验证据

以我国沪深A股中央国有上市公司2012—2020年的数据为样本，将审计署实施的审计监管事件作为一个准自然实验，利用多时点双重差分模型实证检验政府审计对国有企业并购和重大资产重组中业绩承诺可靠性的影响和作用机制。研究发现：（1）政府审计的实施能够显著提升国有企业并购重组业绩承诺的可靠性水平，这一结论在全部并购重组样本中和剔除业绩达标在5%以内的样本后均得到了有效验证，且经过多种稳健性和内生性检验后依然成立；（2）进一步的机制检验结果表明，政府审计能够通过“监督”效应和“增量”效应促进国企并购重组业绩承诺可靠性的提升，但并没有对未接受审计的同行业企业起到很好的“溢出”效应；（3）在业绩承诺到期后的经济后果上，政府审计显著提升了业绩承诺达标企业的长期绩效表现，同时降低了其商誉减值风险，但对于业绩承诺未达标企业的影响效果并不明显。     

Forecasting small area populations with long short-term memory networks

Local and state governments depend on small area population forecasts to make important decisions concerning the development of local infrastructure and services. Despite their importance, current methods often produce highly inaccurate forecasts. Recent years have witnessed promising developments in time series forecasting using Machine Learning across a wide range of social and economic variables. However, limited work has been undertaken to investigate the potential application of Machine Learning methods in demography, particularly for small area population forecasting. In this paper we describe the development of two Long-Short Term Memory network architectures for small area populations. We employ the Keras Tuner to select layer unit numbers, vary the window width of input data, and apply a double training and validation regime which supports work with short time series and prioritises later sequence values for forecasts. These methods are transferable and can be applied to other data sets. Retrospective small area population forecasts for Australia were created for the periods 2006–16 and 2011–16. Model performance was evaluated against actual data and two benchmark methods (LIN/EXP and CSP-VSG). We also evaluated the impact of constraining small area population forecasts to an independent national forecast. Forecast accuracy was influenced by jump-off year, constraining, area size, and remoteness. The LIN/EXP model was the best performing method for the 2011-based forecasts whilst deep learning methods performed best for the 2006-based forecasts, including significant improvements in the accuracy of 10 year forecasts. However, benchmark methods were consistently more accurate for more remote areas and for those with populations <5000.     

Brand placements in video games: How local in-game experiences influence brand attitudes

Brand placements are omnipresent in video games, but their overall effect on brand attitudes is small and varies substantially between studies. The present research takes an evaluative conditioning perspective to explain when and how brand placements in video games influence brand attitudes. In two experiments with a 3D first-person video game, we show that only brands encountered during positive in-game experiences benefit from the placement, but not those encountered during negative in-game experiences. Building on the cognitive processes underlying evaluative conditioning, we also show that brand attitudes largely depend on the memory for the pairing of a brand with positive/negative in-game experiences. Pairing memory and thus also evaluative conditioning effects increase when players attend to the pairing of brands and positive/negative experiences, for example, when such pairings are a central part of the game's storyline. Overall, our findings show that evaluative conditioning and its cognitive mechanisms can be utilized to explain and predict advertising effects in applied settings, such as brand placements in video games.     

基于Nand Flash的星载综合数据固态记录系统

在星载平台资源受限条件下，采用以FPGA+CPU为控制核心、Nand Flash为固态存储阵列的系统架构，实现了高速、大容量、高可靠的数据记录。针对传统双Plane操作与并行扩展对存储速度提升有限、芯片使用较多的问题，采用4级流水线方式控制Flash阵列。为解决标准传输协议传输效率低的问题，设计了一种自定义高速串行传输协议。为减缓空间辐射环境对存储数据的影响，采用了三模冗余、配置回读与部分重构等容错机制。对所提出系统进行的实验验证结果表明，该星载记录系统存储容量达36 Tbit，记录与回放速度分别达到16 Gbit/s与8 Gbit/s，传输误码率为10-12，传输包效率为96.7%，可作为通用存储系统以满足航天应用需求。     

How tourism memory boosts creativity? The role of openness to experience

Although tourism memory is widely regarded as a valuable commodity, scant research examines its influence on creativity, a crucial factor for individual and organizational success. Moreover, while both travel and work are vital for human functioning, there are inconsistent views on whether they are opposing domains. This research aims to fill these gaps by investigating the unique characteristics of tourism memories and their effect on creativity at the individual level. Across four studies, we demonstrate that merely recalling one's tourism memory - even from a long time ago - increases creativity. We have also found that this effect is mediated by an increased state of openness to experience triggered by tourism memory retrieval, and one's trait openness moderates this effect. Our research contributes to the literature on tourism, management, and psychology, also offering insights for marketers and organizations to leverage the potential of tourism memories to foster creativity and success.