Decision Tree Tool Research Articles

Using Decision Tree Methodology to Predict Employment After Moderate to Severe Traumatic Brain Injury.

Objective:To build decision tree prediction models for long-term employment outcomes of individuals after moderate to severe closed traumatic brain injury (TBI) and assess model accuracy in an independent sample.Setting:TBI Model Systems Centers.Participants:TBI Model Systems National Database participants injured between January 1997 and January 2017 with moderate to severe closed TBI. Sample sizes were 7867 (year 1 postinjury), 6783 (year 2 postinjury), and 4927 (year 5 postinjury).Design:Cross-sectional analyses using flexible classification tree methodology and validation using an independent subset of TBI Model Systems National Database participants.Main Measures:Competitive employment at 1, 2, and 5 years postinjury.Results:In the final employment prediction models, posttraumatic amnesia duration was the most important predictor of employment in each outcome year. Additional variables consistently contributing were age, preinjury education, productivity, and occupational category. Generally, individuals spending fewer days in posttraumatic amnesia, who were competitively employed preinjury, and more highly educated had better outcomes. Predictability in test data sets ranged from a C-statistic of 0.72 (year 5; confidence interval: 0.68-0.76) to 0.77 (year 1; confidence interval: 0.74-0.80).Conclusion:An easy-to-use decision tree tool was created to provide prognostic information on long-term competitive employment outcomes in individuals with moderate to severe closed TBI. Length of posttraumatic amnesia, a clinical marker of injury severity, and preinjury education and employment status were the most important predictors.

Simplifying wound dressing selection for residential aged care

To demonstrate the effectiveness of a simplified wound dressing regimen and develop a decision tree tool for wound management, for use by nurses in a residential aged care setting. A three-phase process was used. Firstly, practice was audited, including a cost analysis of existing wound treatment. Secondly, staff were educated on wounds, wound care products and a simplified wound management protocol. In the final phase, the new wound care products and protocol used for skin tears and other wounds (as assessed by a nurse practitioner) were evaluated and ongoing education provided as required. A total of 93 residents from two residential aged care facilities participated in the study. Overall, 178 wounds were identified, of which 121 were considered eligible for treatment under the new protocol. The majority of residents were aged >85 years and the major wound type was skin tears 72.7% (n=88). The mean healing time for skin tears was significantly shorter than for other wounds (19.7±14.2 days versus 30.9±25.2 days, p=0.0359). Most wounds were suitable for the simplified dressing selection regimen. Data collected from a survey of nursing staff demonstrated that the simplified protocol products were easy to apply, conformed well, were easy to remove and stayed in place. In addition, patient satisfaction was positive and removal of dressings was generally pain free. In summary, data collected from this project showed that the dressings were effective, performed well and simplified wound dressing selection for nursing staff.

Clinically relevant improvements achieved from a facilitated implementation of a gestational diabetes model of care.

Medical nutrition therapy is a cornerstone treatment in gestational diabetes; however, most Australian women diagnosed with gestational diabetes do not receive this. The project evaluated adaptation of a successful evidence-based gestational diabetes model of care implementation from a tertiary centre into regional sites with varied demographics, population size and service capacity. The project used a hub (project team)-spoke (sites) model in Far North Queensland (Site 1) and regional South-East Queensland (Site 2). Sites selected demonstrated strong gestational diabetes team cohesiveness and project commitment. The project phases were consultation, baseline, transition and implementation. A best practice decision tree tool was provided to assess/manage barriers to the model of care and clinical outcomes captured through a project database. Role clarification of site members, management engagement, site visits, decision tree and database refinement were completed in the project's first phase. Unexpected organisational and team barriers prevented timeline implementation as planned. Sites negotiated relevant reallocation of resources to achieve project deliverables. The proportion of women seen according to best practice increased from 3.5 to 87.8% (P < 0.001) (Site 1) and nil to 4.8% (P = 0.09) (Site 2), and those on medication dropped by 3.4 (Site 1) and 9.1% (Site 2). This project demonstrates a successful implementation using a facilitated and rigorous approach. Support, engagement and tools at many levels were keys to success at both sites. The present study illustrates the opportunities and challenges of conducting implementation research within routine clinical care, particularly in resource-challenged sites.

Simple decision-tree tool to facilitate author identification of reporting guidelines during submission: a before\u2013after study

BackgroundThere is evidence that direct journal endorsement of reporting guidelines can lead to important improvements in the quality and reliability of the published research. However, over the last 20 years, there has been a proliferation of reporting guidelines for different study designs, making it impractical for a journal to explicitly endorse them all. The objective of this study was to investigate whether a decision tree tool made available during the submission process facilitates author identification of the relevant reporting guideline.MethodsThis was a prospective 14-week before–after study across four speciality medical research journals. During the submission process, authors were prompted to follow the relevant reporting guideline from the EQUATOR Network and asked to confirm that they followed the guideline (‘before’). After 7 weeks, this prompt was updated to include a direct link to the decision-tree tool and an additional prompt for those authors who stated that ‘no guidelines were applicable’ (‘after’). For each article submitted, the authors’ response, what guideline they followed (if any) and what reporting guideline they should have followed (including none relevant) were recorded.ResultsOverall, 590 manuscripts were included in this analysis—300 in the before cohort and 290 in the after. There were relevant reporting guidelines for 75% of manuscripts in each group; STROBE was the most commonly applicable reporting guideline, relevant for 35% (n = 106) and 37% (n = 106) of manuscripts, respectively. Use of the tool was associated with an 8.4% improvement in the number of authors correctly identifying the relevant reporting guideline for their study (p < 0.0001), a 14% reduction in the number of authors incorrectly stating that there were no relevant reporting guidelines (p < 0.0001), and a 1.7% reduction in authors choosing a guideline (p = 0.10). However, the ‘after’ cohort also saw a significant increase in the number of authors stating that there were relevant reporting guidelines for their study, but not specifying which (34 vs 29%; p = 0.04).ConclusionThis study suggests that use of a decision-tree tool during submission of a manuscript is associated with improved author identification of the relevant reporting guidelines for their study type; however, the majority of authors still failed to correctly identify the relevant guidelines.

A decision tree tool supporting the assessment of groundwater vulnerability

The Water Framework Directive and Groundwater Directive aim at preserving and improving the groundwater status. Groundwater bodies are classified as being or not being at risk of failing to meet these objectives. Those at risk are subject to more precise risk assessment where the concept of vulnerability is considered in the pathway part of the source–pathway–receptor scheme. However, no further details on implementation strategies are provided. In order to support groundwater managers and decision-makers in implementation of programs protecting groundwater, a systematic operational approach based on a decision tree is proposed, which leads the user through the stages of vulnerability assessment. First, a problem has to be formulated related to a threatening of the quantitative and/or qualitative status of a groundwater body. Next, the stated problem needs to be related to the intrinsic or specific vulnerability. Methods used for the intrinsic vulnerability assessment belong to two categories: subjective rating and objective methods. Method selection depends primarily on: data availability, knowledge and available resources. A key issue is the lag time associated with transport between a source/event of contamination and the water body. This lag time is primarily controlled by the temporal scale of water flow. It provides information about flow processes and at the same time also about timescales required for the implementation of strategies. Effects of any measures taken cannot be observed immediately but at the earliest after these estimated lag times emphasizing the need to also proactively safeguard groundwater resources and preserve their good status.

A data mining approach for turning potential customers into real ones in basket purchase analysis

Today, in this highly competitive, quality-based world attracting customers is very important. In line with is always right slogan, customer relationship management is at the core of an organisation's strategy. It plays an important role in four aspects, including: customer identification, customer attraction, customer retention, and customer satisfaction. By customer's life cycle's analysis, commercial organisations have obtained increased customer value. Data mining and data warehousing tools and other customer relation management methods have provided new opportunities for businesses. This paper discusses the practical use of data mining in identification of potential customers and potential customers' criteria in the competitive business environment. It also presents the mechanism for identification of potential customers who have the potential to become real customers. After analysing customers' basket purchase, based on the identified parameters, a pattern emerges. By using the decision tree tool, we identify the main criteria and its several sub-criteria and determine their importance level. We determine their importance in order to analyse customers' basket purchase and turn potential customers into real ones. Based on the proposed model, we suggest that organisations increase their investment on the customers that have potential to turn into a loyal customer.

Mobile decision-tree tool technology as a means to detect wildlife crimes and build enforcement networks

Accurate field identification of illegally traded wildlife and wildlife products is critically important in the detection and suppression of wildlife crimes. Yet many law enforcement officers and concerned citizens lack access to resources for identifying species and products; this is particularly true for those with no formal expertise in biology, zoology or wildlife training. Emerging digital technologies such as mobile applications may provide important easy-to-use decision-tree style tools for in situ identification. With local government and civil society partners, we are piloting such tools in China and Vietnam to identify whole animals and ivory products; and in the United States developing tools that will be used at U.S. military bases in Afghanistan to identify species from wildlife products. We are coordinating these efforts to minimize redundancy and overhead; we benefit from shared backend support for a photo database and species ID keys that can be translated easily to ensure enough flexibility for targeting needs of the specific country and audience. Planned inclusion of ‘ask the expert’ and geolocation functions will increase accuracy in identification and aid monitoring and research of supply chains. For these emerging technologies to be successful, deployment must be accompanied with on-the-ground trainings to recruit and retain enforcement personnel. The establishment of a supporting network of experts and a user community will be critical for long-term implementation and evaluation of success. Preliminary response from users of a pilot app in China demonstrates high potential for employing these technologies as routine tools to help fight wildlife crimes.

A new approach for discretizing continuous attributes in learning systems

Discretization is a process to convert continuous attributes into discrete format to represent signals for further data processing in learning systems. The main concern in discretization techniques is to find an optimal representation of continuous values with limited number of intervals that can effectively characterize the data and meanwhile minimize information loss. In this paper, we propose a novel class-attribute interdependency discretization algorithm (termed as NCAIC), which takes account of data distribution and the interdependency between all classes and attributes. In our proposed solution, the upper approximation of rough sets as a prime part of the discretization algorithm is applied, and the class-attribute mutual information is used to automatically control and adjust the scope of the discretization of continuous attributes. Some experiments with comparison to five other discretization algorithms are reported, where 13 benchmarked datasets extracted from UCI database and the well-known C4.5 decision tree tool are employed in this study. Results demonstrate that in general our proposed algorithm outperforms other tested discretization algorithms in terms of classification performance.

Decision trees

Decision trees trace their origins to the era of the early development of written records. This history illustrates a major strength of trees: exceptionally interpretable results which have an intuitive tree‐like display which, in turn, enhances understanding and the dissemination of results. The computational origins of decision trees—sometimes called classification trees or regression trees—are models of biological and cognitive processes. This common heritage drives complementary developments of both statistical decision trees and trees designed for machine learning. The unfolding and progressive elucidation of the various features of trees throughout their early history in the late 20th century is discussed along with the important associated reference points and responsible authors. Statistical approaches, such as a hypothesis testing and various resampling approaches, have coevolved along with machine learning implementations. This had resulted in exceptionally adaptable decision tree tools, appropriate for various statistical and machine learning tasks, across various levels of measurement, with varying levels of data quality. Trees are robust in the presence of missing data and offer multiple ways of incorporating missing data in the resulting models. Although trees are powerful, they are also flexible and easy to use methods. This assures the production of high quality results that require few assumptions to deploy. The treatment ends with a discussion of the most current developments which continue to rely on the synergies and cross‐fertilization between statistical and machine learning communities. Current developments with the emergence of multiple trees and the various resampling approaches that are employed are discussed. WIREs Comput Stat 2013, 5:448–455. doi: 10.1002/wics.1278This article is categorized under: Statistical Learning and Exploratory Methods of the Data Sciences &gt; Clustering and Classification Statistical Learning and Exploratory Methods of the Data Sciences &gt; Pattern Recognition Statistical Learning and Exploratory Methods of the Data Sciences &gt; Rule-Based Mining

Southwest Pacific Ocean response to a warmer world: Insights from Marine Isotope Stage 5e

Paleoceanographic archives derived from 17 marine sediment cores reconstruct the response of the Southwest Pacific Ocean to the peak interglacial, Marine Isotope Stage (MIS) 5e (ca. 125 ka). Paleo‐Sea Surface Temperature (SST) estimates were obtained from the Random Forest model—an ensemble decision tree tool—applied to core‐top planktonic foraminiferal faunas calibrated to modern SSTs. The reconstructed geographic pattern of the SST anomaly (maximum SST between 120 and 132 ka minus mean modern SST) seems to indicate how MIS 5e conditions were generally warmer in the Southwest Pacific, especially in the western Tasman Sea where a strengthened East Australian Current (EAC) likely extended subtropical influence to ca. 45°S off Tasmania. In contrast, the eastern Tasman Sea may have had a modest cooling except around 45°S. The observed pattern resembles that developing under the present warming trend in the region. An increase in wind stress curl over the modern South Pacific is hypothesized to have spun‐up the South Pacific Subtropical Gyre, with concurrent increase in subtropical flow in the western boundary currents that include the EAC. However, warmer temperatures along the Subtropical Front and Campbell Plateau to the south suggest that the relative influence of the boundary inflows to eastern New Zealand may have differed in MIS 5e, and these currents may have followed different paths compared to today.

The Potential Impact of Improving Appropriate Treatment for Fever on Malaria and Non-Malarial Febrile Illness Management in Under-5s: A Decision-Tree Modelling Approach

BackgroundAs international funding for malaria programmes plateaus, limited resources must be rationally managed for malaria and non-malarial febrile illnesses (NMFI). Given widespread unnecessary treatment of NMFI with first-line antimalarial Artemisinin Combination Therapies (ACTs), our aim was to estimate the effect of health-systems factors on rates of appropriate treatment for fever and on use of ACTs.MethodsA decision-tree tool was developed to investigate the impact of improving aspects of the fever care-pathway and also evaluate the impact in Tanzania of the revised WHO malaria guidelines advocating diagnostic-led managementResultsModel outputs using baseline parameters suggest 49% malaria cases attending a clinic would receive ACTs (95% Uncertainty Interval:40.6–59.2%) but that 44% (95% UI:35–54.8%) NMFI cases would also receive ACTs. Provision of 100% ACT stock predicted a 28.9% increase in malaria cases treated with ACT, but also an increase in overtreatment of NMFI, with 70% NMFI cases (95% UI:56.4–79.2%) projected to receive ACTs, and thus an overall 13% reduction (95% UI:5–21.6%) in correct management of febrile cases. Modelling increased availability or use of diagnostics had little effect on malaria management outputs, but may significantly reduce NMFI overtreatment. The model predicts the early rollout of revised WHO guidelines in Tanzania may have led to a 35% decrease (95% UI:31.2–39.8%) in NMFI overtreatment, but also a 19.5% reduction (95% UI:11–27.2%), in malaria cases receiving ACTs, due to a potential fourfold decrease in cases that were untested or tested false-negative (42.5% vs.8.9%) and so untreated.DiscussionModelling multi-pronged intervention strategies proved most effective to improve malaria treatment without increasing NMFI overtreatment. As malaria transmission declines, health system interventions must be guided by whether the management priority is an increase in malaria cases receiving ACTs (reducing the treatment gap), reducing ACT waste through unnecessary treatment of NMFI or expanding appropriate treatment of all febrile illness.

Clinical Issues—January 2012

This MonthVenous thromboembolism prophylaxis in pediatric patientsKey words: pediatric patients, venous thromboembolism, prophylaxis.The Centers for Disease Control and Prevention surgical wound classification systemKey words: surgical wound, wound classification, clean wound, clean-contaminated wound, contaminated wound, dirty infected wound.Using the Surgical Wound Classification Decision Tree toolKey words: surgical wound, wound classification, clean-contaminated wound, contaminated wound.Using cotton surgical masksKey words: allergy, cotton surgical masks, fabric surgical masks, disposable surgical masks.Using povidone-iodine solution for surgical skin antisepsis before thyroid proceduresKey words: povidone iodine, thyroid, surgical skin antisepsis.

A Market Analysis and Decision Tree Tool for Response Analysis: Cash, Local Purchase and/or Imported Food Aid? The Decision Tree Tool

This document provides the Decision Tree Tool itself for the analysis of response options - cash, local purchase or imported food aid. The framework is kept simple for ease of communication: a sequence of specific questions to be answered, each matched with data needs and diagnostic tools commonly useful for effectively answering the question at hand. Maxwell, Lentz and Barrett (2007) explain the rationale for the Tool and its constituent questions. Users will want to consult that document for essential background information and for supplemental considerations that fall outside the scope of this tool. The extensive appendices to this document enumerate common secondary data sources (Appendix 1), primary data collection methods (Appendix 2), and offer more detail (and appropriate technical references) on the specifics of individual diagnostic tools (Appendix 3). Users will want to consult the appendices in detail before trying to implement the tool.

A Market Analysis and Decision Tree Tool for Response Analysis: Cash, Local Purchase and/or Imported Food Aid? Background Paper

This paper attempts to build on that framework to develop practical tools for field decision makers, although these decision tools are related specifically to the question of a food access shortfall at the household level, which may be related to a food availability shortfall at market, regional or national level.

Optimisation of the decision tree technique applied to simulated sow herd datasets

In this study, the robustness of the C4.5-decision tree algorithm was applied to sow herd datasets for investigating the limitations of this analysing technique. First, simulated sow herd datasets including inconsistent farmer culling policies which appear in real datasets were classified. These results were compared with very uniform and simple replacement rules. Furthermore, an optimisation of different pruning methods which can be changed in the decision tree tool was done. The evaluation parameters of all classifications were calculated with the stratified fold cross-validation and varying the number of folds showed that 10 folds were an appropriate number of subdividing the datasets. By simplifying the sow selection in the simulation, the sensitivity and error rate of the datasets showed improved values. In particular, datasets with randomly selected and inconsistent culling rules showed less sensitivities between 20 and 53%. A comparison of the classification of datasets with pruning or without pruning showed that with pruning, smaller sizes of trees resulted. The pruning class had a decrease of 23 leaves and 46 nodes, as compared to the without-pruning class, in the highest branching example. Differences exist between the two different pruning methods in the classification parameters and also in the tree size in dependence of the sow herd performance level and the size of the datasets.

A practical decision tree tool that water utilities can use to solve taste and odor problems

Much research has been performed into the science of aesthetic water quality, but utilities have had difficulty incorporating this science into their practices. A decision tree was developed that, for the first time, provides a resource that utility professionals can use to describe an off-flavor, determine the origin, quickly learn how to proceed with an investigation of the off-flavor event and identify potential solutions. Utility personnel can quickly learn how to be detectives for their systems without the training necessary to be experts. If the electronic decision tree does not lead to resolution of the taste and odor problem, then the utility professional, who was not previously knowledgeable in off-flavors, can have a productive discussion with an expert regarding tastes and odors. Communicating the lessons learned by the scientists of aesthetic water quality to the utility professionals in a quick and easy manner will lead to more resolved taste and odor events and better water quality.

Appropriate Admissions to the Appropriate Unit: A Decision Tree Approach

An intermediate care decision tree tool was developed to meet the demand for intermediate care beds. Concurrently, a charging process was developed to support the acuity adaptable model of care, allowing the patient to remain in the same bed from admission to discharge, regardless of level of care required, adjusting nurse-to-patient ratios as acuity changes. Since beginning this pilot, 96% to 100% of the patients admitted to intermediate care from the emergency department met the criteria. Wait time from request to admission was reduced from 5.5 hours to 2.5 hours. A reduction in nursing costs was noted. The average number of patients waiting daily in the emergency department for an intermediate care bed has been reduced by approximately 80%. A significant difference in length of stay was not noted.

A Fuzzy Decision Tree Approach To Start AGenetic Algorithm For Data Classification

This paper introduces a fuzzy decision tree to initiate the first population of a genetic algorithm to perform data classification. On large datasets, the evolutive process tends to waste computational resources until some good individual is found. It is expected that the use of a fuzzy decision tree can significantly reduce this feature. The genetic algorithm aims to obtain small fuzzy classifiers by means of optimization of fuzzy rules bases. It is shown how a fuzzy rules base is generated from a numerical database and how its best subset is found by the genetic algorithm. The classifiers are evaluated in terms of accuracy, cardinality and number of features employed. The results obtained are compared with a known study in the literature and with an academic decision tree tool. The

Data classification by a fuzzy genetic system approach

This paper presents a fuzzy genetic approach to perform data classification. The method aims to obtain small fuzzy classifiers by means of optimization of fuzzy rules bases using a genetic algorithm. It is shown how a fuzzy rules base is generated from a numerical database and how its best subset is found by the genetic algorithm. The classifiers are evaluated in terms of accuracy, cardinality and amounts of features employed. The results obtained are compared to a known study in the literature and with an academic decision tree tool. The method was able to produce small fuzzy classifiers with very good performance.

A workbook for the preservation management off digital materials

The Arts and Humanities Data Service (AHDS) undertook a research project on digital preservation management funded by Re:source between July 1999 and September 2000. The project aimed to provide a decision tree and other tools to assist organisations in managing the digital materials they create and/or acquire. The main deliverable from the project was a workbook produced by Neil Beagrie and Maggie Jones. The final draft of the workbook was sent for peer review and comment. This paper focuses on the major categories of feedback received from this process and the outcomes of the feedback. Finally, the paper briefly outlines what further developments are anticipated.