深度学习方法在作物遥感分类中的应用和挑战

[目的]准确估算作物的面积和分布对粮食安全至关重要。与传统的机器学习方法相比,深度学习具有多种优势,如端到端训练、可迁移性。为有效利用高时空数据进行作物识别提供了新的机遇。已有多种模型被应用于作物分类任务中,针对不同的分类任务,如何有效地选择模型,并对其进行训练和使用已成为关键问题。[方法]文章回顾了利用深度学习模型对作物分类的主要研究。N维卷积神经网络(N-D CNN)(N=1、2、3)和递归神经网络(RNN)已被有效用于作物分类任务。长短期记忆RNN(LSTM RNN)和门控循环单元RNN(GRU RNN)是RNN的变体,解决了随着时间序列增加RNN出现的梯度消失或爆炸问题。此外,还有研究使用CNN和RNN(我们称为RCNN)的混合模型对作物进行分类。该文首先阐述了使用深度学习方法进行作物制图的背景和意义,并介绍了CNN和RNN模型结构。然后回顾了一些典型的研究,包括模型的结构、遥感数据源、数据处理方法和分类精度。最后,总结了使用深度学习方法进行作物分类的挑战以及现有解决方案的局限性。[结果](1)1-D CNN可用于提取时间特征,或时间+光谱特征,分类效果良好;2-D CNN已被广泛应用于单时相数据的空间特征提取,分类精度依赖于数据源;3-D CNN应用较少,但具有很大的潜力,尤其是时间+空间维度的特征提取;(2)相同条件下(架构、数据源、研究区域、类别),LSTM RNN和GRU RNN分类效果通常高于普通RNN,而前两者的效果差距不大,但GRU RNN训练时间较短;(3)CNN+RNN混合模型(RCNN)用RNN比3-D CNN更适合提取时间特征。这主要是由于RNN建立了对序列数据的长期依赖,而3-D CNN卷积核是局部计算的。[结论]通过分析,认为深度学习技术是作物遥感分类的有效工具。此外,与其他模型相比,RCNN,3-D CNN和GUR RNN具有更大的潜力。     

绿色物流视角下生活垃圾回收处理研究文献综述

近些年来,我国众多城市都出现了"垃圾围城"的现象,垃圾回收处理也成为了社会广泛关注的热点,但是众多学者对相关理论的研究成果并没有被系统性的总结。文章将在绿色物流视角下,对生活垃圾回收处理的相关研究进行全面梳理,从居民垃圾分类意识、垃圾回收处理服务网络构建、垃圾回收站点及中转站的选址等三个方面进行综述,最后进行述评和研究展望。     

军事遗产旅游资源分类与价值评估

军事遗产是重要的历史资源,尤其在历史教育、知识输出、国情教育等方面具有重要价值,因而可被视为独特的旅游资源。本文在国内外军事遗产研究综述的基础上,提出军事遗产的分类方法,共计2个大类、7个亚类、28个子类。在梳理国内外研究的基础上结合专家评价,提出军事遗产的旅游价值评估指标体系,包含要素价值、存续特征2个一级指标及游憩价值、历史价值、审美价值、社会价值、独特程度、规模程度、知名度、完整性、适游性、承载力和安全性11个二级指标,并进一步划分为可观赏性等23个三级指标。采用基于专家打分的层次分析法对指标权重赋值,形成军事遗产旅游价值评价的方法体系,并构建军事遗产旅游价值评价对比关系图,以指导资源群中的单体开发梯次。     

Class formation and commodity chains in the making of regional monocultures: Agrarian reform from above and below in Mexico's henequen and cotton zones

This paper analyses the rise and fall of two regional monocultures in Mexico: the henequen zone in the southern state of Yucatán and the cotton‐growing area of La Laguna. Both regions experienced a dramatic expansion of commodity production between 1870 and 1910, but their key crops came to be cultivated under different labour regimes: debt peonage in the case of henequen and wage labour in the case of cotton. The process of class formation that unfolded in each region culminated in the 1930s in different kinds of crises. In Yucatán, a political struggle between hacienda owners and the federal government resulted in an agrarian reform “from above.” In La Laguna, class conflict between rural wageworkers and the landed bourgeoisie forced an agrarian reform “from below.” These previously distinct labour regimes converged in subsequent decades, however, as rural producers became de facto wageworkers on state‐organized and state‐administered production units known as collective ejidos. Ultimately, changes in the global markets for cotton and henequen, combined with the inability of the Mexican state to reconcile the political logic of agrarian clientelism with shifting commodity chain dynamics, resulted in the collapse of these regional monocultures in the late 20th century.     

Producing Localized Commodity Frontiers at the End of Cheap Nature: An Analysis of Eco‐scalar Carbon Fixes and their Consequences

There is no single ‘great’ commodity frontier whose exploitation under current socio‐technical conditions could fuel capital accumulation at the global scale. According to Jason Moore, this represents the ‘end of Cheap Nature’ and signals a terminal crisis for capitalism as we know it. In this article we complicate this assertion by showing how, in the context of global environmental governance frameworks of carbon control, a diverse range of actors situated at multiple scales are intensifying the use of cities and their hinterlands for the production/transgression of localized commodity frontiers. We draw on scholarship on uneven geographical development, state‐led restructuring and eco‐scalar fixes to present two case studies from different segments of the carbon cycle in the global South. The first case demonstrates how the introduction of waste‐to‐energy technology in Delhi facilitated the generation of ‘carbon credits’ while waste matter itself became a commodity. The second discusses attempts by the Brazilian state of Amazonas (Amazônia) aspiring to shift from rainforest exploitation to financialized conservation supported by the ‘green global city’ functions of metropolitan Manaus. These cases demonstrate that although the global carbon‐control regime may enable accumulation, implementation remains speculative, and localized commodity frontiers provoke social resistances that jeopardize their durability.     

Machine learning for financial transaction classification across companies using character-level word embeddings of text fields

An important initial step in accounting is mapping financial transfers to the corresponding accounts. We devised machine-learning-based systems that automate this process. They use word embeddings with character-level features to process transaction texts. When considering 473 companies independently, our approach achieved an average top-1 accuracy of 80.50%, outperforming baselines that exclude the transaction texts or rely on a lexical bag-of-words text representation. We extended the approach to generalizes across companies and even across different corporate sectors. After standardization of the account structures and careful feature engineering, a single classifier trained on 44 companies from 28 sectors achieved a test accuracy of more than 80%. When trained on 43 companies and tested on the remaining one, the system achieved an average performance of 64.62%. This rate increased to nearly 70% when considering only the largest sector.     

Do country risk and financial uncertainty matter for energy commodity futures?

Using an instrumental variable quantile regression technique, this paper assesses whether country risk and financial uncertainty exert an impact on energy commodity futures prices under different commodity conditional return distributions over the period from January 1994 to July 2017. We also discuss whether the correlations change with different dimensions of country risk, that is economic, financial, and political. The results reveal that country risk and financial stress do have a significant impact on energy commodity returns of futures contracts with different maturities, but their direction, intensity, and significance differ, caused by the distinct market situations and divergent channels of country risk.     

On commodity price limits

This paper examines the behavior of futures prices and trader positions around the occurrence of price limits in commodity futures markets. We ask whether limit events are the result of shocks to fundamental volatility or the result of temporary volatility induced by the trading of noncommercial market participants (speculators). We find little evidence that limits events are the result of speculative activity, but instead associated with shocks to fundamentals that lead to persistent price changes. When futures trading halts price discovery migrates to options markets, but option prices provide a biased estimate of subsequent future prices when trading resumes.     

中国旅游景区功能演变与用地分类构想

研究目的:基于中国旅游景区功能演变、用地特征及问题分析,构建旅游景区用地分类体系,以期为旅游景区用地纳入区域土地利用提供理论基础,为旅游景区规划的深度编制提供实践依据。研究方法:通过调研和问卷厘清现状景区用地情况,对比借鉴相关用地分类体系,基于此构建旅游景区用地分类方案。研究结果:分析并阐明了旅游景区的功能演变、用地特征和现状问题,构建了2大类、9中类、28小类的景区用地分类体系,并与《土地利用现状分类》进行衔接。研究结论:建立可衔接且具可操作性的旅游景区用地分类体系,是实现旅游景区健康可持续发展与用地规范化管控的关键。