Conditional Inference Forests Research Articles

Applying Machine Learning Approach to Explore Childhood Circumstances and Self-Rated Health in Old Age - China and the US, 2020-2021.

Childhood circumstances impact senior health, prompting the introduction of machine learning methods to assess their individual and collective contributions to senior health. Using health and retirement study (HRS) and China Health and Retirement Longitudinal Study (CHARLS), we analyzed 2,434 American and 5,612 Chinese participants aged 60 and above. Conditional inference trees and forests were employed to estimate the influence of childhood circumstances on self-rated health (SRH). The conventional method estimated higher inequality of opportunity (IOP) values in both China (0.039, accounting for 22.67% of the total Gini coefficient 0.172) and the US (0.067, accounting for 35.08% of the total Gini coefficient 0.191). In contrast, the conditional inference tree yielded lower estimates (China: 0.022, accounting for 12.79% of 0.172; US: 0.044, accounting for 23.04% of 0.191), as did the forest (China: 0.035, accounting for 20.35% of 0.172; US: 0.054, accounting for 28.27% of 0.191). Childhood health, financial status, and regional differences were key determinants of senior health. The conditional inference forest consistently outperformed others in predictive accuracy, as demonstrated by lower out-of-sample mean squared error (MSE). The findings emphasize the need for early-life interventions to promote health equity in aging populations. Machine learning showcases the potential in identifying contributing factors.

Is a dementia diagnosis predictive for palliative care? ‐ A machine learning approach

AbstractBackgroundPalliative care (PC) aims to provide dignified and self‐determined life for dying individuals. While the integration of PC for tumor diseases in health care is established in case of progressive course and unfavorable prognosis, PC is surprisingly not mentioned in the German S3‐guideline for dementia. PC is mainly used by older people, with proximity to death rather than age being the determining factor. However, more people are dying at older ages. Moreover, as the elderly population increases, the number of dementia patients also increases. The aim of this paper is to examine the utilization of PC at the end of life and to explore whether a dementia diagnosis is among the top factors that predict PC utilization.MethodWe developed predictive models using discrete survival conditional inference trees (ctrees) and discrete survival conditional inference forests (cforests). We used randomly drawn claims data from a German health insurer from 2014 aged 50+ (N = 250.000), and followed them quarterly until 2019. We analyzed the last eight quarters of life of 13,628 individuals who had died in the observation period and used outpatient (OPC) and inpatient (IPC) PC as outcomes. We included medical history, care measures, medications, comorbidities and patient status (diagnosis of dementia or cancer, combinations of both diagnoses, and diagnosis‐free patients) as possible predictors. Dementia was defined by ICD‐10 codes F00‐F03,F05.1,G23.1,G30,G31.0 and G31.82 and cancer by C00‐C26,C30‐C41,C45‐C58,C60‐C97 and D00‐D09. For evaluation we used concordance‐index (C‐Index) and calibration plots. We identified the most important predictors by using a permutation variable importance approach.ResultThe models for OPC and IPC had a discriminatory power of more than 0.7 (C‐Index) on the test data. The calibration plots showed well calibrated ctrees, but the cforests require further tuning (Figure 1). Among the top 20 predictors of OPC were proximity to death as well as cancer and dementia diagnoses, and their combinations. Neither cancer nor dementia diagnoses were important predictors of IPC.ConclusionWe found that a dementia diagnosis has comparably predictive power for OPC as a cancer diagnosis. This issue should be addressed by including appropriate measures in dementia care guidelines.

Integrated Multimodal Analyses of DNA Damage Response and Immune Markers as Predictors of Response in Metastatic Triple-Negative Breast Cancer in the TNT Trial (NCT00532727).

The TNT trial (NCT00532727) showed no evidence of carboplatin superiority over docetaxel in metastatic triple-negative breast cancer (mTNBC), but carboplatin benefit was observed in the germline BRCA1/2 mutation subgroup. Broader response-predictive biomarkers are needed. We explored the predictive ability of DNA damage response (DDR) and immune markers. Tumor-infiltrating lymphocytes were evaluated for 222 of 376 patients. Primary tumors (PT) from 186 TNT participants (13 matched recurrences) were profiled using total RNA sequencing. Four transcriptional DDR-related and 25 immune-related signatures were evaluated. We assessed their association with objective response rate (ORR) and progression-free survival (PFS). Conditional inference forest clustering was applied to integrate multimodal data. The biology of subgroups was characterized by 693 gene expression modules and other markers. Transcriptional DDR-related biomarkers were not predictive of ORR to either treatment overall. Changes from PT to recurrence were demonstrated; in chemotherapy-naïve patients, transcriptional DDR markers separated carboplatin responders from nonresponders (P values = 0.017; 0.046). High immune infiltration was associated with docetaxel ORR (interaction P values < 0.05). Six subgroups were identified; the immune-enriched cluster had preferential docetaxel response [62.5% (D) vs. 29.4% (C); P = 0.016]. The immune-depleted cluster had preferential carboplatin response [8.0% (D) vs. 40.0% (C); P = 0.011]. DDR-related subgroups were too small to assess ORR. High immune features predict docetaxel response, and high DDR signature scores predict carboplatin response in treatment-naïve mTNBC. Integrating multimodal DDR and immune-related markers identifies subgroups with differential treatment sensitivity. Treatment options for patients with immune-low and DDR-proficient tumors remains an outstanding need. Caution is needed using PT-derived transcriptional signatures to direct treatment in mTNBC, particularly DDR-related markers following prior chemotherapy.

Freshwater-related species richness in Natura 2000 sites strongly depends on the surrounding land use besides local habitat conditions

The Birds and Habitats Directive are the cornerstones of Europe's nature conservation policy and the resulting Natura 2000 (N2k) sites form the largest coordinated network of protected areas in the world. Despite the ambitious targets of these directives and decades of efforts, biodiversity, especially of freshwater-related species, continues to decline in Europe. While multiple stressors at larger spatial scales are known to limit the effect of river restoration projects, the importance of surrounding land use outside the N2k sites for freshwater-related species richness inside N2k sites has rarely been studied. Conditional inference forests were used to assess the importance of land use in the surrounding and upstream of the German N2k sites compared to local habitat conditions inside. Freshwater-related species richness depended on land use in the surrounding besides local habitat conditions. Results indicated that this was especially true for birds in small N2k sites embedded in a wet, diverse, and patchy landscape and for non-birds due to the provision of additional habitats outside the N2k sites. Given that most N2k sites in Europe are rather small, the surrounding habitat conditions and land use potentially influences and affects freshwater-related species in many N2k sites across Europe. The additional conservation and restoration areas to be designated under the EU Biodiversity Strategy and upcoming EU restoration law should either be large enough or surrounded by extensive land use to optimize their effect on freshwater-related species.

An identified agronomic interpretation for potassium permanganate oxidizable carbon

AbstractThe absence of clear empirical relationships between soil health and agronomic outcomes remains an obstacle to widespread adoption of soil health assessments in row crop systems. The objectives of this research were to (1) determine whether soil health indicators are connected to corn (Zea mays L.) productivity and (2) establish interpretive benchmarks for soil health indicators in Missouri. The objectives were accomplished by collecting corn grain yield at 446 monitoring sites (37 m2) in 84 commercial production fields in 2018–2020. Soil health and soil fertility samples were collected prior to planting at each site. These data, along with site‐specific soil and weather data, were modeled using traditional stepwise regression and nonparametric random forest (RF) and conditional inference forest (CIF) approaches. Root‐mean‐square errors were similar (1.4–1.5 Mg ha−1) with distinct R2 improvements over stepwise regression for both CIF (R2 = 0.45) and RF (R2 = 0.46) algorithms. Only seasonal rainfall and potassium permanganate oxidizable carbon (POXC) were included as top factors governing grain productivity in each model approach, thus demonstrating a regionally robust empirical relationship between POXC and grain productivity. Partial dependency analysis and two decision tree approaches identified 415 mg POXC kg−1 as a threshold for maximum grain productivity, providing a framework for regional interpretation of on‐farm soil health assessments. Little evidence was found connecting grain productivity with autoclaved citrate extractable protein and soil respiration. These findings underscore the power of POXC as an emerging soil health indicator to assess and quantify soil management effects on grain productivity.

Moderators of peak respiratory exchange ratio during exercise testing in children and adolescents with Fontan physiology.

Many patients with Fontan physiology are unable to achieve the minimum criteria for peak effort during cardiopulmonary exercise testing. The purpose of this study is to determine the influence of physical activity and other clinical predictors related to achieving peak exercise criteria, signified by respiratory exchange ratio ≥ 1.1 in youth with Fontan physiology. Secondary analysis of a cross-sectional study of 8-18-year-olds with single ventricle post-Fontan palliation who underwent cardiopulmonary exercise testing (James cycle protocol) and completed a past-year physical activity survey. Bivariate associations were assessed by Wilcoxon rank-sum test and simple regression. Conditional inference forest algorithm was used to classify participants achieving respiratory exchange ratio > 1.1 and to predict peak respiratory exchange ratio. Of the n = 43 participants, 65% were male, mean age was 14.0 ± 2.4 years, and 67.4% (n = 29) achieved respiratory exchange ratio ≥ 1.1. Despite some cardiopulmonary exercise stress test variables achieving statistical significance in bivariate associations with participants achieving respiratory exchange ratio > 1.1, the classification accuracy had area under the precision recall curve of 0.55. All variables together explained 21.4% of the variance in respiratory exchange ratio, with peak oxygen pulse being the most informative. Demographic, physical activity, and cardiopulmonary exercise test measures could not classify meeting peak exercise criteria (respiratory exchange ratio ≥ 1.1) at a satisfactory accuracy. Correlations between respiratory exchange ratio and oxygen pulse suggest the augmentation of stroke volume with exercise may affect the Fontan patient's ability to sustain high-intensity exercise.

Retinopathy prediction in type 2 diabetes: Time-varying Cox proportional hazards and machine learning models

BackgroundDiabetic retinopathy (DR) is one of the most common complications in type 2 diabetes (T2D) with an estimated prevalence of 22%. Predictive modelling has largely been dependent on Cox proportional hazards (CPH) with assumptions of linearity and constant hazards. Machine learning (ML) approaches may prove advantageous in more adequately capturing non-linear effects. ObjectiveTo construct and compare DR prediction models using CPH and ML models with time-varying covariates. DesignReal-world, retrospective cohort study. SettingA tertiary care hospital in Thailand. ParticipantsData on 48,622 T2D patients from electronic health records between 1st January 2010 and 31st December 2019. MethodsTime-to-event time-varying models that included 13 variables were trained in diabetic retinopathy prediction. The CPH and ML models were compared using left-truncated right censoring relative risk forest (LTRC-RRF) and left-truncated right censoring conditional inference forest (LTRC-CIF) algorithms. ResultsThe CPH model outperformed both ML approaches with a Harrell's C-index (c-index) of 0.70 compared to c-indices of 0.51–0.57 for the ML models in the test dataset. Both CPH and ML models showed insulin use and the presence of chronic kidney disease increased DR risk. Sodium glucose transporter 2 inhibitors and dyslipidemia were associated with reduced DR risk. ConclusionCPH provided better predictive power for DR risk than ML modelling using real world data. The presence of comorbidities and the use of antidiabetic medications were associated with the greatest drivers of DR risk.

Optimal Tuning of Random Survival Forest Hyperparameter with an Application to Liver Disease.

Random Forest (RF) is a technique that optimises predictive accuracy by fitting an ensemble of trees to stabilise model estimates. The RF techniques were adapted into survival analysis to model the survival of patients with liver disease in order to identify biomarkers that are highly influential in patient prognostics. The methodology of this study begins by applying the classical Cox proportional hazard (Cox-PH) model and three parametric survival models (exponential, Weibull and lognormal) to the published dataset. The study further applied the supervised learning methods of Tuning Random Survival Forest (TRSF) parameters and the conditional inference Forest (Cforest) to optimally predict patient survival probabilities. The efficiency of these models was compared using the Akaike information criteria (AIC) and integrated Brier score (IBS). The results revealed that the Cox-PH model (AIC = 185.7233) outperforms the three classical models. We further analysed these data to observe the functional relationships that exist between the patient survival function and the covariates using TRSF. The IBS result of the TRFS demonstrated satisfactory performance over other methods. Ultimately, it was observed from the TRSF results that some of the covariates contributed positively and negatively to patient survival prognostics.

Optimal zinc level and uncertainty quantification in agricultural soils via visible near-infrared reflectance and soil chemical properties

Zinc (Zn) is a vital element required by all living creatures for optimal health and ecosystem functioning. Therefore, several researchers have modeled and mapped its occurrence and distribution in soils. Nonetheless, leveraging model predictive performances while coupling information derived from visible near-infrared (Vis-NIR) and soils (i.e. chemical properties) to estimate potential toxic elements (PTEs) like Zn in agricultural soils is largely untapped. This study applies two methods to rapidly monitor Zn concentration in agricultural soil. Firstly, employing Vis-NIR and machine learning algorithms (MLAs) (Context 1) and secondly, applying Vis-NIR, soil chemical properties (SCP), and MLAs (Context 2). For the Vis-NIR information, single and combined pretreatment methods were applied. The following MLAs were used: conditional inference forest (CIF), partial least squares regression (PLSR), M5 tree model (M5), extreme gradient boosting (EGB), and support vector machine regression (SVMR) respectively. For context 1, the results indicated that M5-MSC (M5 tree model-multiplicative scatter correction) with coefficient of determination (R2) = 0.72, root mean square error (RMSE) = 21.08 (mg/kg), median absolute error (MdAE) = 13.69 and ratio of performance to interquartile range (RPIQ) = 1.63 was promising. Regarding context 2, CIF with spectral pretreatment and soil properties [CIF-DWTLOGMSC + SCP (conditional inference forest-discrete wavelet transformation-logarithmic transformation-multiplicative scatter correction-soil chemical properties)] yielded the best performance of R2 = 0.86, RMSE = 14.52 (mg/kg), MdAE = 6.25 and RPIQ = 1.78. Altogether, for contexts 1 and 2, the CIF-DWTLOGMSC + SCP approach (context 2) was the best Zn model outcome for the agricultural soil. The uncertainty map revealed a low to high error distribution in context 1, and a low to moderate distribution in context 2 for all models except CIF, which had some patches with high uncertainty. We conclude that a multiple optimization approach for modeling Zn levels in agricultural soils is invaluable and may provide fast and reliable information needed for area-specific decision-making.

Epi-MEIF: detecting higher order epistatic interactions for complex traits using mixed effect conditional inference forests.

Understanding the relationship between genetic variations and variations in complex and quantitative phenotypes remains an ongoing challenge. While Genome-wide association studies (GWAS) have become a vital tool for identifying single-locus associations, we lack methods for identifying epistatic interactions. In this article, we propose a novel method for higher-order epistasis detection using mixed effect conditional inference forest (epiMEIF). The proposed method is fitted on a group of single nucleotide polymorphisms (SNPs) potentially associated with the phenotype and the tree structure in the forest facilitates the identification of n-way interactions between the SNPs. Additional testing strategies further improve the robustness of the method. We demonstrate its ability to detect true n-way interactions via extensive simulations in both cross-sectional and longitudinal synthetic datasets. This is further illustrated in an application to reveal epistatic interactions from natural variations of cardiac traits in flies (Drosophila). Overall, the method provides a generalized way to identify higher-order interactions from any GWAS data, thereby greatly improving the detection of the genetic architecture underlying complex phenotypes.

Selection of a Transparent Meta-Model Algorithm for Feasibility Analysis Stage of Energy Efficient Building Design: Clustering vs. Tree

Energy Efficient Building (EEB) design decisions that have traditionally been made in the later stages of the design process now often need to be made as early as the feasibility analysis stage. However, at this very early stage, the design frame does not yet provide sufficient details for accurate simulations to be run. In addition, even if the decision-makers consider an exhaustive list of options, the selected design may not be optimal, or carefully considered decisions may later need to be rolled back. At this stage, design exploration is much more important than evaluating the performance of alternatives, thus a more transparent and interpretable design support model is more advantageous for design decision-making. In the present study, we develop an EEB design decision-support model constructed by a transparent meta-model algorithm of simulations that provides reasonable accuracy, whereas most of the literature used opaque algorithms. The conditional inference tree (CIT) algorithm exhibits superior interpretability and reasonable classification accuracy in estimating performance, when compared to other decision trees (classification and regression tree, random forest, and conditional inference forest) and clustering (hierarchical clustering, k-means, self-organizing map, and Gaussian mixture model) algorithms.

Using Conditional Inference Forests to Examine Predictive Ability for Future Falls and Syncope in Older Adults: Results from The Irish Longitudinal Study on Ageing.

The extent to which gait and mobility measures predict falls relative to other risk factors is unclear. This study examined the predictive accuracy of over 70 baseline risk factors, including gait and mobility, for future falls and syncope using conditional inference forest models. Data from 3 waves of The Irish Longitudinal Study on Ageing (TILDA), a population-based study of community-dwelling adults aged ≥50 years were used (n = 4 706). Outcome variables were recurrent falls, injurious falls, unexplained falls, and syncope occurring over 4-year follow-up. The predictive accuracy was calculated using 5-fold cross-validation; as there was a class imbalance, the algorithm was trained using undersampling of the larger class. Classification rate, the area under the receiver operating characteristic curve (AUROC), and area under the precision recall curve (PRAUC) assessed predictive accuracy. Highest overall accuracy was 69.7% for recurrent falls in 50-64-year olds. AUROC and PRAUC were ≤0.69 and ≤0.39, respectively, for all outcomes indicating low predictive accuracy. History of falls, unsteadiness while walking, fear of falling, mobility, medications, mental health, and cardiovascular health and function were the most important predictors for most outcomes. Conditional inference forest models using over 70 risk factors resulted in low predictive accuracy for future recurrent, injurious and unexplained falls, and syncope in community-dwelling adults. Gait and mobility impairments were important predictors of most outcomes but did not discriminate well between fallers and non-fallers. Results highlight the importance of multifactorial risk assessment and intervention and validate key modifiable risk factors for future falls and syncope.

Ensemble methods for survival function estimation with time-varying covariates.

Survival data with time-varying covariates are common in practice. If relevant, they can improve on the estimation of a survival function. However, the traditional survival forests-conditional inference forest, relative risk forest and random survival forest-have accommodated only time-invariant covariates. We generalize the conditional inference and relative risk forests to allow time-varying covariates. We also propose a general framework for estimation of a survival function in the presence of time-varying covariates. We compare their performance with that of the Cox model and transformation forest, adapted here to accommodate time-varying covariates, through a comprehensive simulation study in which the Kaplan-Meier estimate serves as a benchmark, and performance is compared using the integrated difference between the true and estimated survival functions. In general, the performance of the two proposed forests substantially improves over the Kaplan-Meier estimate. Taking into account all other factors, under the proportional hazard setting, the best method is always one of the two proposed forests, while under the non-proportional hazard setting, it is the adapted transformation forest. -fold cross-validation is used as an effective tool to choose between the methods in practice.

UAV Remote Sensing for Detecting within-Field Spatial Variation of Winter Wheat Growth and Links to Soil Properties and Historical Management Practices. A Case Study on Belgian Loamy Soil

Intra-field heterogeneity of soil properties, such as soil organic carbon (SOC), nitrogen (N), phosphorous (P), exchangeable cations, pH, or soil texture, is a function of complex interactions between biological factors, physical factors, and historic agricultural management. Mapping the crop growth and final yield heterogeneity and quantifying their link with soil properties can contribute to an optimization of amendment/fertilizer application and crop yield in a management variable zones (MVZ) approach. To this end, we studied a field of 17 ha consisting of four former fields that were merged in early 2017 and cropped with winter wheat in 2018. Historical management practices data were collected. The topsoil characteristics were analyzed by grid-based sampling and kriged to create maps. We tested the capacity of a multispectral MicaSense® RedEdge-MTM camera sensor embedded on an unmanned aerial vehicle (UAV) to map in-season growth of winter wheat. Relating several vegetation indices (VIs) to the plant area index (PAI) measured in the field highlighted the red-edge NDVI (RENDVI) as the most suitable to follow the crop growth throughout the growing season. The georeferenced final grain yield of the winter wheat was measured by a combine harvester. The spatial patterns in RENDVI at three phenological stages were mapped and analyzed together with the yield map. For each of these images a conditional inference forest (CI-forest) algorithm was used to identify the soil properties significantly influencing these spatial patterns. Historical management practices of the four former fields have induced significant heterogeneity in soil properties and crop growth. The spatial patterns of RENDVI are rather constant over time and their Spearman rank correlation with yield is similar along the growing season (r ≃ 0.7). Soil properties explain between 87% (mid-March) to 78% (mid-May) of the variance in RENDVI throughout the growing season, as well as 66% of the variance in yield. The pH and exchangeable K are the most significant factors explaining from 15 to 26% of the variance in crop growth. The methodology proposed in this paper to quantify the importance of soil parameters based on the CI-forest algorithm can contribute to a better management of amendment/fertilizer inputs by stressing the most important parameters to take into consideration for site-specific management. We also showed that heterogeneity induced by the soil properties can be described by a crop map early in the season and that this crop map can be used to optimize soil sampling and thus amendment/fertilizer management.

VIPPID: a gene-specific single nucleotide variant pathogenicity prediction tool for primary immunodeficiency diseases.

Distinguishing pathogenic variants from non-pathogenic ones remains a major challenge in clinical genetic testing of primary immunodeficiency (PID) patients. Most of the existing mutation pathogenicity prediction tools treat all mutations as homogeneous entities, ignoring the differences in characteristics of different genes, and use the same model for genes in different diseases. In this study, we developed a single nucleotide variant (SNV) pathogenicity prediction tool, Variant Impact Predictor for PIDs (VIPPID; https://mylab.shinyapps.io/VIPPID/), which was tailored for PIDs genes and used a specific model for each of the most prevalent PID known genes. It employed a Conditional Inference Forest model and utilized information of 85 features of SNVs and scores from 20 existing prediction tools. Evaluation of VIPPID showed that it had superior performance (area under the curve = 0.91) over non-specific conventional tools. In addition, we also showed that the gene-specific model outperformed the non-gene-specific models. Our study demonstrated that disease-specific and gene-specific models can improve SNV pathogenicity prediction performance. This observation supports the notion that each feature of mutations in the model can be potentially used, in a new algorithm, to investigate the characteristics and function of the encoded proteins.

Comparison of tree-based methods used in survival data

Abstract Survival trees and forests are popular non-parametric alternatives to parametric and semi-parametric survival models. Conditional inference trees (Ctree) form a non-parametric class of regression trees embedding tree-structured regression models into a well-defined theory of conditional inference procedures. The Ctree is applicable in a varietyof regression-related issues, involving nominal, ordinal, numeric, censored, as well as multivariate response variables and arbitrary measurement scales of covariates. Conditional inference forests (Cforest) consitute a survival forest method which combines a large number of Ctrees. The Cforest provides a unified and flexible framework for ensemble learning in the presence of censoring. The random survival forests (RSF) methodology extends the random forests method enabling the approximation of rich classes of functions while maintaining generalisation errors low. In the present study, the Ctree, Cforest and RSF methods are discussed in detail and the performances of the survival forest methods, namely the Cforest and RSF have been compared with a simulation study. The results of the simulation demonstrate that the RSF method with a log-rank score distinction criteria outperforms the Cforest and the RSF with log-rank distinction criteria.

Development of machine learning prediction models to explore nutrients predictive of cardiovascular disease using Canadian linked population-based data.

Machine learning may improve use of observational data to understand the nutritional epidemiology of cardiovascular disease (CVD) through better modelling of non-linearity, non-additivity, and dietary complexity. Our objective was to develop machine learning prediction models for exploring how nutrients are related to CVD risk and to evaluate their predictive performance. We established a population-based cohort from the Canadian Community Health Survey and measured CVD incidence and mortality from 2004 to 2018 using administrative databases of national hospital discharges and deaths. Predictors included 61 nutrition variables and fourteen socioeconomic, demographic, psychological, and behavioural variables. Conditional inference forest models were interpreted and evaluated by permutation feature importance, accumulated local effects, and predictive discrimination and calibration. A total of 12 130 individuals were included in the study. Use of supplements, caffeine, and alcohol were the most important nutrition variables for prediction of CVD. Supplement use was associated with decreased risk, caffeine was associated with increasing risk, and alcohol had a u-shaped association with risk. The model had an out-of-sample c-statistic of 0.821 (95% confidence interval = 0.801-0.842). Exploratory findings included both known and novel associations and predictive performance was competitive, suggesting that further application of machine learning to nutritional epidemiology may help elucidate risks and improve predictive models. Novelty: Machine learning prediction models were developed for CVD using dietary data. Models were interpreted with interpretable machine learning techniques, revealing diverse associations between diet and CVD. Models achieved comparable or superior predictive performance to existing CVD risk prediction models.

Not all sales performance is created equal: personality and interpersonal traits in inbound and outbound marketing activities.

A long tradition of research has shown that personality traits, such as extraversion and agreeableness, and interpersonal constructs better predict job performance with a tacit but not explicit distinction in sales marketing activities. In this contribution, we aim to understand the role of job-related and interest data, interpersonal, and personality traits in affecting either inbound or outbound marketing activities and the overall sales performance. An original questionnaire integrates the interpersonal traits and personality factors reported in the literature in sales marketing activities (independent variables). The results were matched with the individual job-related and interest data (control variables) and sales performance (criterion variables) - expressed as the total number of closed contracts over the inbound/outbound related contacts of employees with responsibility in marketing activities for a large banking group. We are able to identify the relevant predictors of sales performance by creating full binary trees using control and independent variables in conditional inference forests and variable importance index measures. Higher performers in either inbound or outbound marketing activities rely on distinct personality sub-traits, which have fundamentally essential implications for interpersonal functioning, and personal data when agreeableness is central to the ability to function effectively in the interpersonal realm of sales activity.

Modeling lapse rates using machine learning: a comparison between survival forests and cox proportional hazards techniques

This study undertakes a comparative analysis of the performance of machine learning and traditional survival analysis techniques in the insurance industry. The techniques compared are the traditional Cox Proportional Hazards (CPH), Random Survival Forests (RSF) and Conditional Inference Forests (CIF) machine learning models. These techniques are applied in a case study of insurance portfolio of one of Ecuador's largest insurer. This study demonstrates how machine learning techniques perform better in predicting survival function measured by the C-index and Brier Score. It also demonstrates that the predictive contribution of covariates in the RSF model is consistent with the traditional CPH model.

A comparative study of forest methods for time-to-event data: variable selection and predictive performance

BackgroundAs a hot method in machine learning field, the forests approach is an attractive alternative approach to Cox model. Random survival forests (RSF) methodology is the most popular survival forests method, whereas its drawbacks exist such as a selection bias towards covariates with many possible split points. Conditional inference forests (CIF) methodology is known to reduce the selection bias via a two-step split procedure implementing hypothesis tests as it separates the variable selection and splitting, but its computation costs too much time. Random forests with maximally selected rank statistics (MSR-RF) methodology proposed recently seems to be a great improvement on RSF and CIF.MethodsIn this paper we used simulation study and real data application to compare prediction performances and variable selection performances among three survival forests methods, including RSF, CIF and MSR-RF. To evaluate the performance of variable selection, we combined all simulations to calculate the frequency of ranking top of the variable importance measures of the correct variables, where higher frequency means better selection ability. We used Integrated Brier Score (IBS) and c-index to measure the prediction accuracy of all three methods. The smaller IBS value, the greater the prediction.ResultsSimulations show that three forests methods differ slightly in prediction performance. MSR-RF and RSF might perform better than CIF when there are only continuous or binary variables in the datasets.For variable selection performance,When there are multiple categorical variables in the datasets, the selection frequency of RSF seems to be lowest in most cases. MSR-RF and CIF have higher selection rates, and CIF perform well especially with the interaction term.The fact that correlation degree of the variables has little effect on the selection frequency indicates that three forest methods can handle data with correlation.When there are only continuous variables in the datasets, MSR-RF perform better. When there are only binary variables in the datasets, RSF and MSR-RF have more advantages than CIF.When the variable dimension increases, MSR-RF and RSF seem to be more robustthan CIFConclusionsAll three methods show advantages in prediction performances and variable selection performances under different situations. The recent proposed methodology MSR-RF possess practical value and is well worth popularizing. It is important to identify the appropriate method in real use according to the research aim and the nature of covariates.