A Note on Breiman's Random Forest Data Mining Technique and Conventional Cox Modeling of Survival Statistics: The Case of the Phantom “Induct” Covariate in the Ohio State University Kidney Transplant Database

This note represents a portion of the research that has been conducted using the Ohio State University's large kidney transplant database. Our latest results into understanding the impact of covariates on renal graft success including the impact of drug therapy are discussed.

The major result here is that using both the Cox model and Breiman's Random Forest Data Mining techniques has helped to unravel the mystery of the “induct” immunosuppressant covariate slipping in and out of the list of important variables. We also make the interesting observation that the Random Forest method seems to mimic the clinician's (R.P.) understanding of the importance of variables.     

The soil seed bank and its relationship to the aboveground vegetation in deciduous forests in New York City

The soil seed bank was studied in two deciduous forests in Bronx (New York City), NY. The purpose of this study was to determine how the biotic andabiotic differences between urban and rural forests arereflected in urban forest seed banks. Soil samples werecollected in two consecutive years and monitored for emergencein the greenhouse over two years. In 1993, the mean number ofemergents ranged from 4636 to 5373 m^-2 (excluding ferns), or from 6972 to 9651 m^-2 (including ferns). In 1994, the mean number of emergents ranged from 1656 to 2013 m^-2 (excluding ferns), or from 5019 to 5992 m^-2 (including ferns). Graminoids and fernscombined accounted for approximately 70% of all emergents eachyear. Three taxa, Rubus spp., Betula lenta, and Liriodendron tulipifera, comprised 60–80% of the woody emergents and were theonly woody taxa to exhibit delayed germination. A substantialnumber of forbs, graminoids and ferns (15–50%) exhibiteddelayed germination. The nonnative woody species Ailanthusaltissima, Morus alba, and Celastrus orbiculatus were absentfrom the aboveground vegetation of some forest plots yet werepresent at low densities in the seed bank. A greater meandensity of emergents and the presence of nonnative species arethe main differences between the seed banks of these urbanforests and those reported for similar nonurban forests in thisregion.     

An integrated approach to evaluating urban forest functionality

Despite the fact that forests in urban areas play multiple and often conflicting roles, research and management efforts are typically geared towards a single role or purpose. Urban ecology addresses this multiplicity of function by viewing human and natural systems in urban areas not as separate entities, but as interacting components of an integrated whole. We present an interdisciplinary approach for evaluating the different ways that forests are often valued: economically, socially, and ecologically in residential areas of King County, WA. Economic function is measured as the change in housing prices attributed to location on the gradient, using a hedonic price model. For social function we use a survey to measure (1) residents’ use of parks and forests, and (2) satisfaction with their neighborhoods. We measure ecological function as songbird species richness, using bird survey data. Overlaying the curves of economic, social, and ecological function on the common axis of our urban gradient allows for relationships and tradeoffs to be qualitatively evaluated. Each function responds differently to the gradient. The housing price response is strongest at high and low levels of urbanization, with positive premiums in both areas. Satisfaction with neighborhood attributes decreases with increasing urbanization, while the likelihood of mentioning ‘parks’ as an important element of a resident’s neighborhood increases. Songbird richness peaks in less-developed areas. Evaluating the different functions together is an important step in recognizing and understanding the multiple roles forested areas play.     

Wildfire Exposure Analysis on the National Forests in the Pacific Northwest,USA

We analyzed wildfire exposure for key social and ecological features on the national forests in Oregon and Washington. The forests contain numerous urban interfaces, old growth forests, recreational sites, and habitat for rare and endangered species. Many of these resources are threatened by wildfire, especially in the east Cascade Mountains fire‐prone forests. The study illustrates the application of wildfire simulation for risk assessment where the major threat is from large and rare naturally ignited fires, versus many previous studies that have focused on risk driven by frequent and small fires from anthropogenic ignitions. Wildfire simulation modeling was used to characterize potential wildfire behavior in terms of annual burn probability and flame length. Spatial data on selected social and ecological features were obtained from Forest Service GIS databases and elsewhere. The potential wildfire behavior was then summarized for each spatial location of each resource. The analysis suggested strong spatial variation in both burn probability and conditional flame length for many of the features examined, including biodiversity, urban interfaces, and infrastructure. We propose that the spatial patterns in modeled wildfire behavior could be used to improve existing prioritization of fuel management and wildfire preparedness activities within the Pacific Northwest region.     

产业集群中社会资本创新绩效研究

本文分析了集群中社会资本的形成机制、演变特征及其对集群创新绩效的影响.作者认为,集群中社会资本的形成机制及演化路径是决定集群创新绩效的关键因素,而要保持集群持续的创新优势.就必须转变社会资本积累的传统方式,使社会资本积累突破既有的传统文化和地缘关系的限制.     

An analytical study of the classification of highly skewed data

This article proposes a discriminant function and an algorithm to analyze the data addressing the situation, where the data are positively skewed. The performance of the suggested algorithm based on the suggested discriminant function (LNDF) has been compared with the conventional linear discriminant function (LDF) and quadratic discriminant function (QDF) as well as with the nonparametric support vector machine (SVM) and the Random Forests (RFs) classifiers, using real and simulated datasets. A maximum reduction of approximately 81% in the error rates as compared to QDF for ten-variate data was noted. The overall results are indicative of better performance of the proposed discriminant function under certain circumstances.     

西南少数民族林业谚语的生态思想解析

林业谚语是千百年来人们处理林人关系的经验总结。西南少数民族中传承着丰富的林业谚语,是西南少数民族利用森林与认识森林的思想结晶,其间隐含着用林—护林—用林的林业生态逻辑,其内核是林人共生思想。分析西南少数民族林业谚语,探寻西南少数民族与森林相处的思想轨迹,从中探取森林可持续利用的秘诀,可为处理现代社会因森林遭到毁坏而导致的生态问题提供借鉴和参考。     

关于退耕还林还草若干问题的思考

该文主要针对目前退耕还林还草工程在规划设计、政策补助、基层单位工作经费及工作负荷等方面存在的实际问题进行分析,并提出建议：对退耕还林还草年度计划的制订、下达时间进行调整;坚持生态效益优先的规划原则,慎重规划经济林（兼用林）;延长补助年限,并尽快以政策形式确定下来;适当提高种苗费补助标准;关注基层林业单位的工作经费和工作负荷。     

Bayesian credit ratings: A random forest alternative approach

Cerciello and Giudici (2014 Cerciello, P., Giudici, P. (2014). Bayesian credit ratings. Commun. Stat. Theory Methods. 43:867–878.[Taylor &; Francis Online], [Web of Science ®] , [Google Scholar]) proposed a Bayesian approach to improve the ordinal variable selection in credit rating assessment. However, no comparison has been made with other methods and the predictive power was not tested. This study proposes an integrated framework of random forest (RF)-based methods and Bayesian model averaging (BMA) to validate and investigate the ordinal variable importance in evaluating credit risk and predicting default in greater depth. The proposed approach was superior to the Cerciello and Giudici method in terms of predictive accuracy and interpretability when applied to a European credit risk database.