Proposed Decision Tree Research Articles

A practical guide to prescribing sublingual immunotherapy tablets in North America for pediatric allergic rhinoconjunctivitis: an injection-free allergy immunotherapy option.

Allergic rhinoconjunctivitis (ARC) is a common disease that affects individuals of all ages. Pediatricians may be the first (and only) point of care for children with ARC. Sublingual immunotherapy (SLIT)-tablets are a convenient at-home, injection-free allergy immunotherapy option that can be used for the treatment of ARC. This paper provides a practical guide for pediatricians to aid in prescribing SLIT-tablets to children with ARC in North America. Topics include a summary of the available SLIT-tablets and their efficacy and safety, guidance on when SLIT-tablets are an appropriate option, and how to diagnose ARC and identify culprit allergens. Practical guidance is also provided through a proposed decision tree, a prescribing checklist and prescribing procedures, and suggested follow-up assessments.

최빈값 추정을 위한 의사결정나무

Decision trees are one of the data mining techniques that make predictions by recursively partitioning data structures based on split rules. Since the analysis results can be understood through the tree structure, it has the advantage of having high interpretation power as well as predictive power. In addition, it is used in many fields because it is able to identify nonlinear relationships between response and predictor variables. However, if the purpose of it is to predict the mode of the response variable, there is a limitation in that the previously proposed decision tree cannot be applied. Thus, we develop a new form of the modal decision tree model by integrating the kernel density estimation methods into the decision tree model. The simulation is conducted with four models. The results are compared for each size of the data when the predictor variable and the response variable are linear and nonlinear relationship cases. When the data has a linear relationship, the performance of the modal desicion tree model proposed in this paper is comparative to that of the previously proposed modal linear regression (MODLR) model. When the data has a nonlinear relationship, the performance of the modal tree model is better.

Prediction of Crop Yield Based-on Soil Moisture using Machine Learning Algorithms

Agriculture planning is playing an important role as the economic growth is very high day by day in our daily life. There is lot of research study going on as there are few important issues like soil nutrients, crop prediction, farming system, crop monitoring in agriculture with modern farming system. Based on various parameters, farming issues and farming system, there is lot of change in production rate and market prices. Crop prediction and crop monitoring is main factor to produce good quality of crops for farmers to predict crop yield based on soil moisture. Prediction of crop yield includes forecasting factors like temperature, humidity, rainfall, etc., and crop yield based on soil moisture includes few measures like pH, NPK (Nitrogen, Phosphorous and potassium) values using various sensors. Farmers can predict or come to a decision the type of soil moisture values, farmers can decide the type of crop to be planted. In this paper, we proposed decision tree supervised machine learning algorithm to improve our results for the prediction of crop yield based on soil moisture parameters to achieve economic growth for achievement of better results.

Developing guidelines for waste designation of biofuel feedstocks in carbon footprints and life cycle assessment

Biofuels continue to play a role in strategies for decarbonization, especially for transportation. However, to determine whether a particular biofuel pathway can achieve decarbonization, its life cycle greenhouse gases (i.e., carbon intensity) must be calculated. The two most important steps in the biofuel life cycle are feedstock production and conversion. When a feedstock is derived from a waste or residue, it can lead to very low carbon intensity for the resulting fuel. Consistency in how waste or residue designations are made is critical, especially when carbon intensities are used in a regulatory context. We develop a taxonomy for feedstock classification, supported by a decision tree for the designation of feedstocks, which may be classified as primary products, co-products, by-products, or wastes. These designations are critical for determining the allocation method by which carbon intensity estimates for feedstocks are calculated under policies such as California's Low Carbon Fuel Standard. The decision tree is designed to include two possible questions as the fifth step, differentiated by the inclusion or exclusion of an economic threshold. An economic threshold is common for distinguishing between co-products and byproducts in biofuel policies. Based on the resulting decision tree, several feedstock candidates are tested, and the results analyzed through a SWOT (Strengths, Weaknesses, Opportunities, Threats) analysis to determine if there are specific benefits or drawbacks of the proposed decision tree. The results from the feedstock tests show that the choice of the fifth question does not affect carbon intensity estimates for the selected test cases. However, depending on the economic value of a feedstock, a feedstock could be deemed a byproduct instead of a co-product based on an economic threshold. Choosing to include an economic threshold in the decision process will result in more feedstocks designated as by-products, rather than primary products, thus attributing them a lower carbon intensity.

PREDICTION OF PLASTIC-TYPE FOR SORTING SYSTEM USING DECISION TREE MODEL

Plastic is the most widely used inorganic material globally, but its hundred-year disintegration time can harm the environment. Polyethylene Terephthalate (PET/PETE), High-Density Polyethylene (HDPE), and Polypropylene are all commonly used plastics that have the potential to become waste (PP). An essential first step in the recycling process is sorting out plastic waste. A low-cost automated plastic sorting system can be developed by using digital image data in the red, green, and blue (RGB) color space as the dataset and predicting the type using learning datasets. This paper proposes the Decision Tree model to predict the three plastic-type sorting systems based on discretizing predictor variables into two and three categories. The resampling method of k-fold cross-validation with ten folds for less biased. Discretization of the predictor variables into three categories informs that the proposed decision tree model has higher performance compared to the two categories with an accuracy of 81.93 %, a recall-micro of 72.89 %, a recall-macro of 72.30 %, a specificity-micro of 86.45%, and the specificity-macro of 86.51%, respectively. The micro is determined by the number of decisions made for each object. In comparison, the macro is calculated based on the average decision made by each class.

Improvement of transformer dissolved gas analysis interpretation using j48 decision tree model

<span lang="EN-US">Dissolved gas analysis (DGA) is widely accepted as an effective method to detect incipient faults within power transformers. Gases such as hydrogen, methane, acetylene, ethylene and ethane are normally utilized to identify the transformer fault conditions. Several techniques have been developed to interpret DGA results such as the key gas method, Doernenburg, Rogers, IEC ratio-based methods, Duval Triangles, and the latest Duval Pentagon methods. However, each of these approaches depends on the experts' shared knowledge and experience rather than quantitative scientific methods, therefore different diagnoses may be reported for the same oil sample. To overcome these shortcomings, this paper proposed the use of decision tree method to interpret the transformer health condition based on DGA results. The proposed decision tree model employed three main fault gases; methane, acetylene, ethylene as inputs, and classified the transformer into eight fault conditions. The J48 algorithm is used to train and developed the decision tree model. The performance of the proposed model is validated with the pre-known condition of transformers and compared with the Duval Triangle method. Results show that the proposed model delivers better precision and accuracy in predicting transformer fault conditions compared to DTM with 81% and 69% respectively.</span>

4D Trajectory Conflict Detection and Resolution Using Decision Tree Pruning Method

The aviation community develops Trajectory Based Operations (TBO) as an advancement in Air Traffic Management (ATM). There is still the need for an efficient scheme to present the trajectories, manage their associated data, and further detect and resolve the conflicts (CDR) that should eventually occur. In this research, we develop a CDR framework for managing predicted 4-Dimensional Trajectory (4DT). Using Not Only SQL (NoSQL) database (Cassandra and MongoDB), the 4D trajectories of related routes are presented, and the possible conflicts are detected using the strategy of Computing in NoSQL Database. Compared with other conflict detection algorithms, usually by the pairwise method with O(n2) at least, the proposed Decision Tree Pruning Method (DTPM) effectively treats massive data sets. The 4DT data are collected by Trajectory Predictor (TP) concerning 58% of the whole Brazilian air traffic. The comparison results between Cassandra and MongoDB from the case studies show the effectiveness of the proposed methods for conflict detection. In addition, we prove that the conflict resolution approach is viable for application in real scenarios, finding near-optimal solutions for the conflicts identified by the framework. Finally, we also demonstrated the development of sustainable artificial intelligence in intelligent air transportation to improve safety in air traffic management.

A decision tree network with semi-supervised entropy learning strategy for spectroscopy aided detection of blood hemoglobin

Non-invasive techniques for rapid blood testing are gaining traction in global healthcare as they optimize medical screening, diagnosis and clinical decisions. Fourier transform infrared (FT-IR) spectroscopy is one of the most common technologies that can be used for non-destructive aided medical detection. Typically, after acquiring the Fourier transform infrared spectrum, spectral data preprocessing and feature extraction and quantitative analysis of several indicators of blood samples can be accomplished, in combination with chemometric method studies. At present, blood hemoglobin (HGB) concentration is one of the most valuable information for the clinical diagnosis of patient’s health status. FT-IR spectroscopy is employed as a green technique aided medical test of blood HGB. Then the acquired HGB concentration data is switched to the spectral feature data by the studies of advanced chemometric method, in help for hiding the sensitive medical information to protect the privacy of patients. The decision tree network architecture is proposed for feature extraction of FT-IR data in order to find the small set of wavenumbers that are able to quantify HGB. A semi-supervised learning strategy is designed for tuning the number of network neuron nodes, in the way of searching for the maximum entropy increment. Each neuron is optimized by the growing of a semi-supervised decision tree, to accurately identify the informative FT-IR wavenumbers. The features extracted by the semi-supervised learning decision tree network guarantees the FT-IR aided detection model has high efficiency and high prediction accuracy. A model of quantifying the HGB concentration shows that the proposed decision tree network with semi-supervised entropy learning strategy outperforms the usual methods of full spectrum partial least square model and the fully connected neural network model in prediction accuracy. The framework is expected to support the FT-IR spectral technology for aided detection of medical and clinical data.

A Data-Driven Predictive Model for Speed Control in Automotive Safety Applications

The importance of transportation to society and daily life makes safe driving an important responsibility to reduce the occurrence of accidents. Among the accident factors, overspeeding led to a maximum probability of road crashes and death. As the number of vehicles increases, manually monitoring the overspeed offense is laborious and time-consuming. The existing context-aware-based speed control strategy is effective but pricey. Therefore, in this article, a cost-effective, compact, and simple design-oriented speed recommendation and control mechanism is designed to avoid accidents caused by excessive speed. The proposed system leverages the potential of the decision tree regression model to do data analysis and find out the recommended speed for a particular location at a specific time. In the implementation process, the trained regression model is stored in the Heroku web server to obtain the speed. To support the recommendation, model of a new dataset is created from the Gray Spot dataset and stored in the Storage bucket of Firebase (Open Source), which comprises the time, location, weather, and speed attributes. The vehicle user receives speed recommendations from the Heroku server for every location; consequently, the speed controller regulates the speed of the motor through a motor driver’s electrical signal. The proposed decision tree regressor achieved 92.3% prediction accuracy during processing with 30 locations. The efficiency of the model showed that this model is a good assistant to car drivers and an optimal solution to reduce accidents on a larger level.

A Machine Learning-Based Approach for Classification of Focal Splenic Lesions Based on Their CT Features.

The aim of the study was to describe the CT features of focal splenic lesions (FSLs) in dogs in order to predict lesion histotype. Dogs that underwent a CT scan and had a FSL diagnosis by cytology or histopathology were retrospectively included in the study. For the statistical analysis the cases were divided into four groups, based on the results of cytopatholoy or hystopathology, namely: nodular hyperplasia (NH), other benign lesions (OBLs), sarcoma (SA), round cell tumour (RCT). Several qualitative and quantitative CT features were described for each case. The relationship occurring between each individual CT feature and the histopathological groups was explred by means of c chi-square test for the count data and by means of Kruskal-Wallis or ANOVA for the continuous data. Furthermore, the main features of each group were described using factorial discriminant analysis, and a decision tree for lesion classification was then developed. Sarcomas were characterised by large dimensions, a cystic appearance and an overall low post contrast-enhancement. NH and OBLs were characterised by small dimensions, a solid appearance and a high post-contrast enhancement. OBLs showed higher post-contrast values than NH. Lastly, RCTs did not exhibit any distinctive CT features. The proposed decision tree had a high accuracy for the classification of SA (0.89) and a moderate accuracy for the classification of OBLs and NH (0.79), whereas it was unable to classify RCTs. The results of the factorial analysis and the proposed decision tree could help the clinician in classifying FSLs based on their CT features. A definitive FSL diagnosis can only be obtained by microscopic examination of the spleen.

Bogie Stability Control and Management Using Data Driven Analysis Techniques for High-Speed Trains

The bogie is a critical component of a train set. Any failure in the bogies’ structure/mechanism transmission will result in a derailment in the revenue operation. In order to prevent derailment due to bogie failures, the Bogie Instability Detection System (BIDS) is installed on high-speed trains, which provides alarms when abnormality is detected so that appropriate action can be taken. Taking the Taiwan High Speed Rail Corporation (THSRC) as an example, there have been ongoing reports of BIDS alarms since its launch. In almost every case, intermittent warnings are generated when wheel/rail interface faults occur, but no hard fault is found in the bogie’s structure or mechanism. While these reports do not constitute hard faults, corresponding operation actions and inspections are necessary nonetheless. As a result, delays and additional maintenance costs are incurred. In order to save maintenance costs and reduce delays, a decision tree algorithmic approach is proposed to study the potential factors of wheel/rail interface concerns, discover relationships between potential factors, and set up rules for BIDS alarms. In this case study, we found that temperature is the main factor for BIDS alarms; train speed, accumulated mileage, and location are subordinate factors. The proposed decision tree approach can lead to some improvements in preventive maintenance. It also suggests embedding such an intelligent algorithm into BIDS control units for drivers to avoid BIDS alarms.

Acute appendicitis in children: Reexamining indications for conservative treatment – A large prospective analysis

IntroductionConservative antibiotic treatment (CAT) for uncomplicated acute appendicitis (AUA) in children has been proven safe and efficacious. However, as data accumulate, high rates of recurrent appendicitis and subsequent appendectomy have been reported. This prospective longitudinal study evaluated risk-factors for recurrent AUA after CAT in a large cohort, with long-term follow-up. Materials and methodsChildren ages 5 to 16 years admitted to the Department of Pediatric Surgery from 2014 through 2018, diagnosed with AUA were eligible for CAT. We recorded their age, appendix outer diameter, white blood cell counts, C-reactive protein and other related signs and symptoms associated with AUA. Clinical and ultrasonographic follow-up was carried out until follow-up data collection ceased according to the study design (2014–2019). ResultsThe cohort included 646 children who were initially treated successfully with CAT. Among them, 180 (28%) were readmitted for recurrent acute appendicitis during the follow-up period and 138 (21%) eventually had appendectomy. Overall success of 79% for CAT was recorded in this cohort. A multivariable model including; age, sex, appendiceal diameter, WBC and CRP, found the factors of older age, larger outer appendiceal diameter and high WBC counts significantly related to appendectomy during the follow-up period. We offer a decision tree model to predict appendectomy probabilities for patients based on their prognostic measurements. ConclusionCAT in AUA in children should consider older age, larger outer appendiceal diameter and high WBC counts as risk-factors for recurrent AUA and subsequent appendectomy. The proposed decision tree model may help both clinicians and parents before CAT is chosen. Level of evidenceLevel 2

Energy Management Model for HVAC Control Supported by Reinforcement Learning

Heating, ventilating, and air conditioning (HVAC) units account for a significant consumption share in buildings, namely office buildings. Therefore, this paper addresses the possibility of having an intelligent and more cost-effective solution for the management of HVAC units in office buildings. The method applied in this paper divides the addressed problem into three steps: (i) the continuous acquisition of data provided by an open-source building energy management systems, (ii) the proposed learning and predictive model able to predict if users will be working in a given location, and (iii) the proposed decision model to manage the HVAC units according to the prediction of users, current environmental context, and current energy prices. The results show that the proposed predictive model was able to achieve a 93.8% accuracy and that the proposed decision tree enabled the maintenance of users’ comfort. The results demonstrate that the proposed solution is able to run in real-time in a real office building, making it a possible solution for smart buildings.

Data mining technologies in a distributed medical information system

Proposed a method for solving the problem of identifying hidden relationships in hard-to-structure data that have an implicit character is considered using information mining. Proposed decision trees, the effectiveness of which is illustrated by a specific example. The use of OLAP analysis systems on data presented using in the form of a real or virtual hypercube’s information is an effective tool for the effectiveness of the management for medical monitoring

Assessment of potential human health risks in aquatic products based on the heavy metal hazard decision tree

BackgroundNaturally existing and human-produced heavy metals are released into the environment and cannot be completely decomposed by microorganisms, but they continue to accumulate in water and sediments, causing organisms to be exposed to heavy metals.ResultsThis study designs and proposes heavy metal hazard decision trees for aquatic products, which are divided into seven categories including pelagic fishes, inshore fishes, other fishes, crustaceans, shellfish, cephalopods, and algae. Based on these classifications, representative fresh and processed seafood products are at the root of the heavy metal hazard decision trees. This study uses 2,107 cases of eating 556 cooked fresh or processed seafood product samples. The constructions of the proposed decision trees consist of 12 heavy metals, which include inorganic arsenic (iAs), cadmium (Cd), cobalt (Co), chromium (Cr), copper (Cu), iron (Fe), manganese (Mn), nickel (Ni), lead (Pb), strontium (Sr), thallium (Tl), and zinc (Zn). The heavy metal concentrations in cooked fresh and processed seafood product samples are subjected to a food safety risk assessment.ConclusionsThe results indicate the relationships among the seven categories of aquatic products, the relationships among 12 heavy metals in aquatic products, and the relationships among potential human health risks. Finally, the proposed heavy metal hazard decision trees for aquatic products can be used as a reference model for researchers and engineers.

Failure mode prediction of reinforced concrete columns using machine learning methods

In this article, new efficient methods are presented to classify failure modes in reinforced concrete columns. For this purpose, machine learning techniques were utilized with consideration of laboratory datasets collected from the literature. Two different approaches, including decision tree and artificial neural network, have been studied to determine the failure mode of the columns. The variables used to estimate the failure mode were compressive strength of the concrete, span-to-depth ratio, axial load ratio, longitudinal reinforcement ratio, volumetric transverse reinforcement ratio, yield stress of longitudinal reinforcement, and yield stress of transverse reinforcement. A comparison study between the two introduced models indicated that the proposed decision tree provides a desirable accuracy and could specify the failure mode, with no need to a complex calculation. The proposed model has many applications in structural engineering such as seismic evaluation, retrofitting, and rehabilitation as a suitable tool for estimating the failure modes in reinforced concrete columns.

Employment of CHAID and CRT decision tree algorithms to develop bid/no-bid decision-making models for contractors

PurposeBid/no-bid decision is a significant and strategic decision, which must be finalized at an early stage of the bidding process. Such decision-making may have significant impact on the performance of the contractors. Using Chi-square Automatic Interaction Detector (CHAID) and Classification and Regression (CRT) decision tree algorithms, this paper aims to develop bid/no-bid models for design-bid-build projects for contractors.Design/methodology/approachThe models in this study have been developed using CHAID and CRT algorithms. Thirty-four bid/no-bid key factors were collected via extensive research. The bid/no-bid factors were listed based on their importance index as a result of a questionnaire distributed among the construction professionals. These factors were divided into five main risk categories – owner, project, bidding situation, contract and contractor – which were taken as inputs for the models. Split-sample validation was applied for testing and measuring the accuracy of the CHAID and CRT models. Moreover, Spearman's rank correlation and Analysis of Variance (ANOVA) tests were employed to identify the statistical features of the received 169 responses.FindingsThe key bid/no-bid factors in construction industry were categorized in five related groups and ranked based on the relative importance index. It was found that the top 6 ranked bid/no-bid factors were (1) current workload, (2) need for work, (3) previous experience with employer; (4) timely payment by the employer; (5) availability of other projects for bidding (6) reputation of employer in the industry. Matrix comparison between all bid/no-bid groups was performed using Spearman's correlation to measure the relationship between each of the two paired groups. It was concluded that all the relationships were positive.Originality/valueExisting bidding models require many inputs and advanced understanding of mathematics and software to run the model. Contractors tend to use easy, fast and available support methods. Excluding a great number of the bid/no-bid factors may affect the final decision. This paper proposes a bid/no-bid decision tree models for contractors of different sizes. It is the first study in the literature, to the best of authors' knowledge, to study bid/no-bid decision with the proposed decision tree algorithm. The proposed models in this study overcome the shortfalls of most previous models such as avoiding the complexity and difficulties of applying the concept. The proposed model will provide the contractors with a bid/no-bid decision based on the input for the defined bid factor groups. The proposed models display the soft spots and hot spots between the independent and dependent variables, which leads to a better decision. The proposed models display the result effectively in visual terms, easy to understand and easy to apply. The proposed models are a form of multiple effect (or variable) analysis which allows the companies to explain, describe, predict or classify an outcome.

Decision Tree Application to Classification Problems with Boosting Algorithm

A personal credit evaluation algorithm is proposed by the design of a decision tree with a boosting algorithm, and the classification is carried out. By comparison with the conventional decision tree algorithm, it is shown that the boosting algorithm acts to speed up the processing time. The Classification and Regression Tree (CART) algorithm with the boosting algorithm showed 90.95% accuracy, slightly higher than without boosting, 90.31%. To avoid overfitting of the model on the training set due to unreasonable data set division, we consider cross-validation and illustrate the results with simulation; hypermeters of the model have been applied and the model fitting effect is verified. The proposed decision tree model is fitted optimally with the help of a confusion matrix. In this paper, relevant evaluation indicators are also introduced to evaluate the performance of the proposed model. For the comparison with the conventional methods, accuracy rate, error rate, precision, recall, etc. are also illustrated; we comprehensively evaluate the model performance based on the model accuracy after the 10-fold cross-validation. The results show that the boosting algorithm improves the performance of the model in accuracy and precision when CART is applied, but the model fitting time takes much longer, around 2 min. With the obtained result, it is verified that the performance of the decision tree model is improved under the boosting algorithm. At the same time, we test the performance of the proposed verification model with model fitting, and it could be applied to the prediction model for customers’ decisions on subscription to the fixed deposit business.

Quantification of pain severity using EEG-based functional connectivity

Background and objectivesThe traditional pain measures are qualitative and inaccurate. Therefore, electroencephalography (EEG) signals have been recently used and analyzed to differentiate pain from no-pain state. The challenge is emerged when the accuracy of these classifiers is not enough for differentiating between different pain levels. In this paper, we demonstrate that EEG-based functional connectivity graph is remarkably changed by increasing the pain intensity and therefore, by deriving informative features from this graph at each pain level, we finely differentiate between five levels of pain. MethodsIn this research, 23 subjects (mean age: 22 years, Std: 1.4) are voluntarily enrolled and their EEG signals are recorded via 29 electrodes, while they execute the cold pressure test. The signals are recorded two times from each case, while subjects press a button to annotate the EEGs into five pain levels. After denoising the EEGs, the brain connectivity graph in the Alpha band are estimated using partial directed coherence method in successive time frames. By observing the differences of connectivity graph features in different levels of pain, a bio-inspired decision tree (multilayer support vector machines) is proposed. Discriminant features are selected using sequential forward feature selection manner and the selected features are applied to the proposed decision tree. ResultsClassification result for differentiating between the pain and no-pain states provides 92% accuracy (94% sensitivity and 91% specificity), while for the five classes of pain, the proposed scheme generates 86% accuracy (90% sensitivity and 82% specificity), which is slightly decreased compared to the two-class condition. Moreover, the results are evaluated in terms of robustness against noise in different signal to noise ratio levels. Comparison results with previous research imply the significant superiority of the proposed scheme. ConclusionIn this paper, we show that the elicited features from the filtered brain graph are able to significantly discriminate five different levels of pain. This is therefore the amount of co-activation between the brain regions (graph links) are significantly varied, as the pain feeling increases. Our observations are consistent with the physiological observations acquired from the images of functional magnetic resonance and magnetoencephalography.

Trends in Catastrophic Occupational Incidents among Electrical Contractors, 2007-2013.

This study used methodologies of descriptive and quantitative statistics to identify the contributing factors most affecting occupational accident outcomes among electrical contracting enterprises, given an accident occurred. Accident reports were collected from the Occupational Safety and Health Administration’s fatality and catastrophe database. To ensure the reliability of the data, the team manually codified more than 600 incidents through a comprehensive content analysis using injury-classification standards. Inclusive of both fatal and non-fatal injuries, the results showed that most accidents happened in nonresidential buildings, new construction, and small projects (i.e., $50,000 or less). The main source of injuries manifested in parts and materials (46%), followed by tools, instruments, and equipment (19%), and structure and surfaces (16%). The most frequent types of injuries were fractures (31%), electrocutions (27%), and electrical burns (14%); the main injured body parts were upper extremities (25%), head (23%), and body system (18%). Among non-fatal cases, falls (37%), exposure to electricity (36%), and contact with objects (19%) caused most injuries; among fatal cases, exposure to electricity was the leading cause of death (50%), followed by falls (28%) and contact with objects (19%). The analysis also investigated the impact of several accident factors on the degree of injuries and found significant effects from such factors such as project type, source of injury, cause of injury, injured part of body, nature of injury, and event type. In other words, the statistical probability of a fatal accident—given an accident occurrence—changes significantly based on the degree of these factors. The results of this study, as depicted in the proposed decision tree model, revealed that the most important factor for predicting the nature of injury (electrical or non-electrical) is: whether the source of injury is parts and materials; followed by whether the source of injury is tools, instruments, and equipment. In other words, in predicting (with a 94.31% accuracy) the nature of injury as electrical or non-electrical, whether the source of injury is parts and materials and whether the source of injury is tools, instruments, and equipment are very important. Seven decision rules were derived from the proposed decision tree model. Beyond these outcomes, the described methodology contributes to the accident-analysis body of knowledge by providing a framework for codifying data from accident reports to facilitate future analysis and modeling attempts to subsequently mitigate more injuries in other fields.