Voluntary Carbon Reporting Prediction: A Machine Learning Approach

In this paper we address the impact of the introduction of the National Greenhouse and Energy Reporting scheme on corporate carbon reporting, and subsequently identify factors that influence the level of voluntary carbon reporting. A review of the literature demonstrates a large number of potential factors have been previously deployed to explain voluntary reporting practices; however, the analytical and empirical methods widely used in the literature have limiting statistical assumptions and confine analysis to a small number of explanatory factors. To address this limitation in prior research we apply advanced machine learning methods, such as gradient boosting machines and random forests, to identify predictive variables through analytical means. We compare the performance of machine learning methods with traditional methods such as logistic regression. We find that machine learning methods significantly outperform logistic regression and provide fundamentally different interpretations of the role and influence of different predictive variables on voluntary carbon reporting. While most variables were not statistically significant in the logit results, a number of key proxies for financial performance, corporate governance, and corporate social responsibility have out-of-sample predictive power of the level of voluntary carbon reporting in the machine learning analysis.     

A two-stage segment and prediction model for mortgage prepayment prediction and management

Predicting the risk of mortgage prepayments has been the focus of many studies over the past three decades. Most of these works have used single prediction models, such as logistic regressions and survival models, to seek the key influencing factors. From the point of view of customer relationship management (CRM), a two-stage model (i.e., the segment and prediction model) is proposed for analyzing the risk of mortgage prepayment in this research. In the first stage, random forests are used to segment mortgagors into different groups; then, a proportional hazard model is constructed to predict the prepayment time of the mortgagors in the second stage. The results indicate that the two-stage model predicts mortgage prepayment more accurately than the single-stage model (non-segmentation model).     

国际森林问题的分歧及其诠释

以国际森林政策对话机制为研究对象,分析了国际森林问题的内涵和特点,存在的分歧和分歧的实质。国际森林问题分为森林本身问题、森林相关的环境问题和社会问题、导致森林本身问题的成因及解决问题的办法四个层次,超越了森林经营和林业行业,并将国际和国内林业政策对话有机结合起来。国际社会在森林本身存在的问题上有较多的共识,然而发展中国家和发达国家在对问题成因的认知上分歧很大,而就主权、资金援助和技术转让等解决森林问题的手段等分歧更加明显。其分歧体现了参与各方的立场、视角和利益,客观反映了当下尚不具备达成全球法律约束力森林公约的条件。     

保护天然林资源 为后代积累生态资产──伊春林业建设调研报告

运用历史回顾和国际比较的方法,通过对伊春林业建设案例的研究,得出伊春悲剧的关键在于市场失灵和政府失灵,天然林应该得到休养生息,人工林种植不能代替天然林的保护,政府应加大财政支持,同时也应注重市场运作产业升级。     

Water scarcity and food trade in the Southern and Eastern Mediterranean countries

This study investigates food trade patterns in relation to water resources availability in the Southern and Eastern Mediterranean countries (SEMED). Examinations show that most of these countries have a high dependence on the import of water intensive crops – cereal, vegetable oil and sugar, in the domestic food supply. The region as a whole is marginally a net exporter of fruits and vegetables, while variations are substantial across countries. Multi-variable regression analyses show that intensification of water scarcity is an important factor in explaining the increase in food import in the SEMED countries during the past two decades. It also finds that while GDP per capita has a strong influence on the level of food import in a country, its impact on changes in the import during the same period is rather modest. No significant relationship is found between the trade of fruits and vegetables and water resources availability. The projection on food import with respect to the decline in per capita water resources availability results in an increase of 40%, 39% and 14%, respectively, for cereal, vegetable oil and sugar by 2020 in the region, holding other factors constant. The European Union (EU) is the major food trade partner of the SEMED countries, except for cereal. About 70% of the fruit export and 55% of the vegetable export of the region currently go to the EU market. Expanding the export of fruits and vegetables is conducive to improving the value of water use in the SEMED countries. However, the expansion is constrained partly by the barriers in the destination markets, notably the EU.     

Loss of water availability and stream biodiversity under land abandonment and climate change in a Mediterranean catchment (Olzinelles,NE Spain)

In the north rim of the Mediterranean region, where forest cover is increasing as a result of land abandonment and temperatures are rising as a result of climate change, there is increasing interest for the effects of such changes on the runoff of water courses. This is a paramount issue for the conservation of many freshwater habitats and species. In this work we studied the effects of both an increase in forest cover after depopulation and land abandonment and an increase in temperature on the runoff of a Mediterranean catchment and on the aquatic and semi-aquatic fauna species of the stream (Olzinelles valley, NE Spain). Although in our simulation no decreasing trend in runoff is detected, the monthly runoff-rainfall ratio is now 15% lower than 30 years ago, a fact that may be attributed to a drier period rather than to the small afforestation experienced by the catchment in the last decades. Other factors such as increasing temperatures, changing rainfall patterns and increasing canopy cover are discussed. The observed decrease in the water flow has caused the disappearance of white-clawed crayfish (Austropotamobius pallipes), Mediterranean barbel (Barbus meridionalis), chub (Squalius cephalus), European eel (Anguilla anguilla), and southern water vole (Arvicola sapidus). Our results suggest that in a progressively warmer climate, and especially after land abandonment processes, management of Mediterranean mountain areas should be oriented towards an appropriate distribution of agrarian and forest land-covers in terms of water availability. Down to the stream scale, the pools that keep water throughout the year should be conserved and extended to enhance its potential to maintain aquatic and semi-aquatic species populations.     

A graphically based machine learning approach to predict secondary schools performance in Tunisia

The main purpose of this paper is to identify the key factors that impact schools' academic performance and to explore their relationships through a two-stage analysis based on a sample of Tunisian secondary schools. In the first stage, we use the Directional Distance Function approach (DDF) to deal with undesirable outputs. The DDF is estimated using Data Envelopment Analysis method (DEA). In the second stage we apply machine-learning approaches (regression trees and random forests) to identify and visualize variables that are associated with a high school performance. The data is extracted from the Program for International Student Assessment (PISA) 2012 survey. The first stage analysis shows that almost 22% of Tunisian schools are efficient and that they could improve their students’ educational performance by 15.6% while using the same level of resources. Regression trees findings indicate that the most important factors associated with higher performance are school size, competition, class size, parental pressure and proportion of girls. Only, school location appears with no impact on school efficiency. Random forests algorithm outcomes display that proportion of girls at school and school size have the most powerful impact on the predictive accuracy of our model and hence could more influence school efficiency. The findings disclose also the high non-linearity of the relationships between these key factors and school performance and reveal the importance of modeling their interactions in influencing efficiency scores.     

Mixed forests and finance — Methodological approaches

Mixed forest ecosystems show effects of risk compensation similar to diversified portfolios of stocks. Other than portfolios of stocks, the composition of long-lived forests cannot be adjusted at short notice to current market developments. Thus, the financial robustness of tree-species portfolios is very important. However, it is still an open question, which financial approaches are appropriate to evaluate the financial robustness of mixed forests compared to single-species forests.This paper analysed various possibilities to assess portfolios of tree species, namely, the mean-variance (MV) approach applied for the selection of optimal portfolios, the second order stochastic dominance (SSD) criterion and the information-gap approach. For the numerical analysis existing financial data on Norway spruce (Picea abies [L.] Karst.) and European beech (Fagus sylvatica L.) were used which were generated by Monte-Carlo simulations (MCS) under risk.Regardless of the analysed approach, in all cases pure forests did not outperform mixed forests. The MV approach was restricted by the assumption of normally distributed financial returns. It had the tendency to favour larger proportions of the high-risk species Norway spruce. SSD was only able to distinguish between the efficient and the inefficient portfolios. It showed that forests with less than 20% (financial returns generated by MCS) or less than 30% (normally distributed financial returns) of Norway spruce were inefficient. Under information-gap modelling the advantages of risk compensation in mixed forests became clearest. Under the given assumptions, the financial robustness of a mixed forest (50% European beech and 50% Norway spruce) was between 4 and 18 times greater than that of pure Norway spruce.Modern information-gap modelling is very similar to the well known safety first rules and as such it is not that novel to financial optimisation as it seems at the first glance. Although the information-gap theory resulted in the same rankings as the minimisation of the estimated probability of failure, it nevertheless proved to be useful as a supplement to the other methods. Against the background of severe uncertainty inherent in long-term decision-making in forestry, information-gap modelling is based on appropriate assumptions. It favours a decision, which promises the greatest tolerance to possible errors in the available information, while still guaranteeing a minimal acceptable financial result. In principle, classical safety first rules do the same. However, the information-gap model is able to quantify the immunity of a decision against errors in the estimated financial data. From a theoretical point of view, this approach provides a supplementary and methodologically sound opportunity to incorporate severe uncertainty in decision-making.