Reluctant Generalised Additive Modelling

Sparse generalised additive models (GAMs) are an extension of sparse generalised linear models that allow a model's prediction to vary non-linearly with an input variable. This enables the data analyst build more accurate models, especially when the linearity assumption is known to be a poor approximation of reality. Motivated by reluctant interaction modelling, we propose a multi-stage algorithm, called reluctant generalised additive modelling (RGAM), that can fit sparse GAMs at scale. It is guided by the principle that, if all else is equal, one should prefer a linear feature over a non-linear feature. Unlike existing methods for sparse GAMs, RGAM can be extended easily to binary, count and survival data. We demonstrate the method's effectiveness on real and simulated examples.     

Policy lessons from spatiotemporal enrollment patterns of payment for ecosystem service programs in Argentina

Payment for ecosystem services schemes (PES) are lauded as a market-based solution to curtail deforestation and restore degraded ecosystems. However, PES programs often fail to conserve sites under strong long-term deforestation pressures. Underperformance, in part, is likely due to adverse selection. Spatial adverse selection occurs when landowners are more likely to enroll parcels with low deforestation pressure than parcels with high deforestation pressure. Temporal adverse selection arises when parcels are enrolled for short time periods. In both cases, financial resources are allocated without having a sizeable impact on long-term land use change. Improving program performance to overcome these shortcomings requires understanding attributes of landowners and their parcels across large scales to identify spatial and temporal enrollment patterns that drive adverse selection. In this paper, we examine these patterns in Argentina’s PES program in Chaco forest, a global deforestation hotspot. Our study area covers 252,319 km². Results from multinomial logistic regression models showed that large parcels of enrolled land and parcels owned by absentee landowners exhibit greater evidence of spatiotemporal adverse selection than smaller parcels or parcels owned by local landowners. Furthermore, parcels managed under land use plans for conservation and restoration are more likely to be associated with adverse selection than parcels managed for financial returns such as harvest of non-timber forest products, silviculture, and silvopasture. However, prior to recommending that PES programs focus on land uses with higher potential earnings, a greater understanding is needed of the degree to which these land uses meet ecological and biodiversity goals of PES programs. We suggest that increased spatial targeting of enrollment, along with enrollment of local landowners and further incentives for land uses that support conservation and restoration, could promote long-term conservation of forest lands.     

人工智能技术辅助开发“智铸”系统

采用开源免费的深度学习平台,利用python二次开发了“智铸”系统,包括“识铸”“听铸”“盯铸”三个模块,铸造企业员工可以用手机或电脑自主学习,提升员工从业能力。在设备旁安装“听铸”软硬件一体化设备,可以实现对连续运行设备的在线自动监听侦测。“盯铸”模块实现连续生产铸件的自动检测、铸造缺陷分析与重点安全区域的监控。     

The international market selection of Chinese SMEs: How institutional influence overrides psychic distance

The purpose of this research is to contribute to the ongoing debate about whether psychic distance still plays a vital role in the internationalisation of SMEs from emerging markets. Drawing on the prior research which suggests the salient impact of institutional factors on internationalisation, we investigate the role of home country institutions in international market selection. Adopting a multi-case methodology, we collected semi-structured interview data from six small and medium-sized manufacturing firms in China. Our findings suggest that while psychic distance is still important in some circumstances, both formal institutions, such as government support, and informal institutions, such as business and political guanxi, enable Chinese SMEs to choose psychically distant markets. Our findings also indicate that informal institutions interact with formal institutions to further influence SMEs’ international market selection. This research contributes to SME internationalisation studies by revealing how formal and informal institutional factors override psychic distance in influencing international market selection.     

Correction to: Optimal and robust combination of forecasts via constrained optimization and shrinkage

We provide a correction to Proposition 1 in Optimal and robust combination of forecasts via constrained optimization and shrinkage, published in the International Journal of Forecasting 38(1):97-116 (2021). This correction has no impact on any other result (neither theoretical nor empirical) provided in the above paper.     

Non-linear exchange rate relationships: An automated model selection approach with indicator saturation

This paper examines whether the explanatory power of exchange rate models can be improved by allowing for cross-country asymmetries and non-linear effects of fundamentals. Both appear to be crucial. The samples include the USD versus pound and yen from 1982:10 to 2013:10, and automated model selection is conducted with indicator saturation. Several non-linear effects are significant at 1%. Further, many of the indicators present in the linear models are eliminated once allowing for non-linearities; suggesting some of the structural breaks found in previous work were an artifact of the misspecified linear functional form. These conclusions are robust to estimation using principal components.     

How much do alternative cookstoves reduce biomass fuel use? Evidence from North India

Despite widespread global efforts to promote clean cookstoves to achieve improvements in air and forest quality, and to reduce global climate change, surprisingly little is known about the degree to which these actually reduce biomass fuel consumption in real-world settings. Using data from in-house weighing of fuel conducted in rural India, we examine the impact of cleaner cookstoves ⿿ most of which are LPG stoves ⿿ on three key outcomes related to solid fuel use. Our results suggest that using a clean cookstove is associated with daily reductions of about 4.5 kg of biomass fuel, 160 fewer minutes cooking on traditional stoves, and 105 fewer minutes collecting biomass fuels. These findings of substantial savings are robust to the use of estimators with varying levels of control for selection, and to alternative data obtained from household self-reports. Our results support the idea that efforts to promote clean stoves among poor rural households can reduce solid fuel use and cooking time, and that rebound effects toward greater amounts of cooking on multiple stoves are not sufficient to eliminate these gains. We also find, however, that households who have greater wealth, fewer members, are in less marginalized groups, and practice other health-averting behaviors, are more likely to use these cleaner stoves, which suggests that socio-economic status plays an important role in determining who benefits from such technologies. Future efforts to capture social benefits must therefore consider how to promote the use of alternative technologies by poor households, given that these households are least likely to own clean stoves.     

Implementing efficient conservation portfolio design

Modern tools for cost-effective conservation reserve site planning require the planner to have information about spatial distributions of conservation costs and benefits. Climate change creates unprecedented uncertainty about future land values and species habitat ranges, such that conservation scientists cannot map costs and benefits with certainty anymore. This paper contributes to the literature on the economics of conservation in the face of climate change uncertainty. It advances a new method for using modern portfolio theory to choose lands to protect that yield total conservation returns with less uncertainty. It explores the implications for portfolio recommendations of variation in the correlations between ecological and land-value responses to climate change. It also tests the robustness of the method to shortcuts that might be taken to simplify analysis, identifying problems that arise if conservation costs are ignored in portfolio analysis and demonstrating when portfolio recommendations are sensitive to how ecological benefits are quantified.     

大学生速度赛马骑师选材指标体系的构建

通过文献资料、实地考察、专家访问等方法对我国大学生速度赛马骑师选材指标体系进行研究;为保证大学生速度赛马骑师成才的比例,促进其职业的快速发展,需要在选材时从大学生的身体形态、身体素质、身体机能、运动心理等多方面入手,科学合理地挑选出优秀的速度赛马骑师。速度赛马骑师选材指标的构建可以为赛马业大学生骑师选材提供科学指导,增加选材的确定性和可靠性。     

A genetic algorithm based approach to transactional and QoS-aware service selection

As there are various risks of failure when Web services are deployed in unreliable environment, the execution of a composite Web service (CWS) requires the transaction mechanism to guarantee its reliable execution. However, the existing service selection approaches consider QoS and transaction separately and have not considered that transactional properties may affect the QoS such as the execution time of a CWS. This work addresses the importance of considering transactional properties and how the transactional properties affect the QoS simultaneously in the process of service selection by a QoS-aware and transactional-aware selection approach. First, a performance evaluation method is proposed to calculate the execution time of a transactional CWS. Then, a genetic algorithm based approach, which takes into account the execution time, price, transactional property and successful execution rate of CWS, is presented to achieve global optimisation service selection. Finally, experimental results show the efficiency and effectiveness of the approach.