Continuous Attributes Research Articles

Community detection in attributed collaboration network for statisticians

Scientific collaboration helps to promote the dissemination of knowledge and is essential in breeding innovation. Collaboration network analysis is a useful tool to study researchers' collaborations. In this work, we collect papers published between 2001 and 2018 in 43 statistical journals and investigate the collaborative trends and patterns. We find that more and more researchers take part in statistical research, and cooperation among them is strengthening. We further construct an attributed collaboration network and extract its core. Community detection is conducted on the core network by using the edge cross‐validation (ECV) method and the attributed network clustering algorithm (ANCA). In particular, we extend the ANCA to deal with networks having both categorical and continuous attributes. Influential researchers are identified in each community. Furthermore, two kinds of homophily are revealed in our collaboration network: research topic homophily and spatial proximity homophily. Based on the homophily and transitivity, we can make recommendations for researchers. Finally, we compare ANCA with the other three methods and confirm that the combination of nodal attributes and network structure improves the quality of community detection. Our studies show the features of the collaboration among statisticians and present a new perspective to explore researchers.

HT-Fed-GAN: Federated Generative Model for Decentralized Tabular Data Synthesis.

In this paper, we study the problem of privacy-preserving data synthesis (PPDS) for tabular data in a distributed multi-party environment. In a decentralized setting, for PPDS, federated generative models with differential privacy are used by the existing methods. Unfortunately, the existing models apply only to images or text data and not to tabular data. Unlike images, tabular data usually consist of mixed data types (discrete and continuous attributes) and real-world datasets with highly imbalanced data distributions. Existing methods hardly model such scenarios due to the multimodal distributions in the decentralized continuous columns and highly imbalanced categorical attributes of the clients. To solve these problems, we propose a federated generative model for decentralized tabular data synthesis (HT-Fed-GAN). There are three important parts of HT-Fed-GAN: the federated variational Bayesian Gaussian mixture model (Fed-VB-GMM), which is designed to solve the problem of multimodal distributions; federated conditional one-hot encoding with conditional sampling for global categorical attribute representation and rebalancing; and a privacy consumption-based federated conditional GAN for privacy-preserving decentralized data modeling. The experimental results on five real-world datasets show that HT-Fed-GAN obtains the best trade-off between the data utility and privacy level. For the data utility, the tables generated by HT-Fed-GAN are the most statistically similar to the original tables and the evaluation scores show that HT-Fed-GAN outperforms the state-of-the-art model in terms of machine learning tasks.

On the combination of adaptive neuro-fuzzy inference system and deep residual network for improving detection rates on intrusion detection.

Deep Residual Networks (ResNets) are prone to overfitting in problems with uncertainty, such as intrusion detection problems. To alleviate this problem, we proposed a method that combines the Adaptive Neuro-fuzzy Inference System (ANFIS) and the ResNet algorithm. This method can make use of the advantages of both the ANFIS and ResNet, and alleviate the overfitting problem of ResNet. Compared with the original ResNet algorithm, the proposed method provides overlapped intervals of continuous attributes and fuzzy rules to ResNet, improving the fuzziness of ResNet. To evaluate the performance of the proposed method, the proposed method is realized and evaluated on the benchmark NSL-KDD dataset. Also, the performance of the proposed method is compared with the original ResNet algorithm and other deep learning-based and ANFIS-based methods. The experimental results demonstrate that the proposed method is better than that of the original ResNet and other existing methods on various metrics, reaching a 98.88% detection rate and 1.11% false alarm rate on the KDDTrain+ dataset.

An improved decision tree algorithm based on variable precision neighborhood similarity

The decision tree algorithm has been widely used in data mining and machine learning due to its high accuracy, low computational cost and high interpretability. However, when dealing with the continuous data, the classical decision tree algorithm needs to replace continuous attributes with discretized attributes by the strategy of discretization. Discretization may cause a loss of information structure, which will affect the performance of classification. To tackle this problem, many researchers have proposed different decision tree methods based on variable precision neighborhood rough sets. However, these methods do not consider the geometric structure of neighborhood systems, which may lead to a contradiction in the transitivity of the equivalence relation. In this paper, we first define a novel neighborhood geometric similarity in a neighborhood system from the perspective of geometry. Second, by combining the neighborhood geometric similarity and the neighborhood algebraic similarity, we propose four new kinds of neighborhood similarities, which can solve the contradictory transitivity of the equivalence relation. Third, a variable precision neighborhood rough set model is constructed using the new similarities, and a novel decision tree algorithm is proposed based on this model, where the degree of attribute dependence is used as the partition measure. Experimental results on 14 selected datasets from the UCI Machine Learning Repository show that our algorithm is effective. The average accuracy of our algorithm is over 90%, which is 10% higher than the classical decision tree algorithms, and the number of leaf nodes increases slightly.

Adjustment for Unmeasured Spatial Confounding in Settings of Continuous Exposure Conditional on the Binary Exposure Status: Conditional Generalized Propensity Score-Based Spatial Matching

BACKGROUND AND AIM: Generalized propensity score (GPS) is used to estimate causal effects of continuous treatments/exposures. The valid estimation relies on the assumption of no unmeasured confounding. Many environmental, demographic, built-environmental, behavioral, and health data can exhibit spatial patterns, raising concerns of unmeasured spatial confounding. Furthermore, researchers are often confronted with settings where continuous exposure is conditional on binary exposure status (e.g., level of a contaminant (continuous) within a specified radius from residence (binary)). Both binary and continuous exposures can commonly exhibit spatial patterns, such that unmeasured spatial confounding in both dimensions may be concerning. We developed a novel GPS matching method for such settings, called conditional GPS (CGPS)-based spatial matching (CGPSsm). METHODS: CGPSsm estimates the average treatment effect in the treated. CGPSsm matches exposed observational units (e.g., exposed participants) to unexposed units by their spatial proximity and GPS integrated with spatial information. GPS is estimated by separately estimating PS for the binary status (exposed vs. unexposed) and CGPS on the binary status. Spatial regression and machine learning can be leveraged to estimate GPS. A motivating example is to investigate the association between refineries with high petroleum production and refining (PPR) and stroke prevalence in the southeastern United States. RESULTS: CGPSsm maintains the salient benefits of propensity score matching and spatial analysis: straightforward assessments of covariate balance and adjustment for unmeasured spatial confounding. Statistical simulations showed that CGPSsm can adjust for unmeasured spatial confounding. Using our example, we found positive association between PPR and stroke prevalence. CONCLUSIONS: CGPSsm has potential in epidemiological studies where exposure has both binary and continuous attributes and unmeasured spatial confounding may be of concern (e.g., accessibility to emergency services, exposure to nearby environmental factors, and surrounding built environments). Our R package, CGPSspatialmatch, is publicly available. KEYWORDS: Causal inference; Generalized propensity score; Spatial confounding; Machine learning

Using the Diagnostic Odds Ratio to Select Patterns to Build an Interpretable Pattern-Based Classifier in a Clinical Domain: Multivariate Sequential Pattern Mining Study.

BackgroundIt is important to exploit all available data on patients in settings such as intensive care burn units (ICBUs), where several variables are recorded over time. It is possible to take advantage of the multivariate patterns that model the evolution of patients to predict their survival. However, pattern discovery algorithms generate a large number of patterns, of which only some are relevant for classification.ObjectiveWe propose to use the diagnostic odds ratio (DOR) to select multivariate sequential patterns used in the classification in a clinical domain, rather than employing frequency properties.MethodsWe used data obtained from the ICBU at the University Hospital of Getafe, where 6 temporal variables for 465 patients were registered every day during 5 days, and to model the evolution of these clinical variables, we used multivariate sequential patterns by applying 2 different discretization methods for the continuous attributes. We compared 4 ways in which to employ the DOR for pattern selection: (1) we used it as a threshold to select patterns with a minimum DOR; (2) we selected patterns whose differential DORs are higher than a threshold with regard to their extensions; (3) we selected patterns whose DOR CIs do not overlap; and (4) we proposed the combination of threshold and nonoverlapping CIs to select the most discriminative patterns. As a baseline, we compared our proposals with Jumping Emerging Patterns, one of the most frequently used techniques for pattern selection that utilizes frequency properties.ResultsWe have compared the number and length of the patterns eventually selected, classification performance, and pattern and model interpretability. We show that discretization has a great impact on the accuracy of the classification model, but that a trade-off must be found between classification accuracy and the physicians’ capacity to interpret the patterns obtained. We have also identified that the experiments combining threshold and nonoverlapping CIs (Option 4) obtained the fewest number of patterns but also with the smallest size, thus implying the loss of an acceptable accuracy with regard to clinician interpretation. The best classification model according to the trade-off is a JRIP classifier with only 5 patterns (20 items) that was built using unsupervised correlation preserving discretization and differential DOR in a beam search for the best pattern. It achieves a specificity of 56.32% and an area under the receiver operating characteristic curve of 0.767.ConclusionsA method for the classification of patients’ survival can benefit from the use of sequential patterns, as these patterns consider knowledge about the temporal evolution of the variables in the case of ICBU. We have proved that the DOR can be used in several ways, and that it is a suitable measure to select discriminative and interpretable quality patterns.

Machine learning and shallow groundwater chemistry to identify geothermal prospects in the Great Basin, USA

This study discovers various geothermal prospects in the Great Basin, USA based on shallow groundwater chemical (geochemical) data. The geochemical data are expected to include hidden (latent) information that is a proxy for geothermal prospectivity. We processed the sparse geochemical data in the Great Basin at 14,341 locations including 18 attributes. Next, a non-negative matrix factorization with customized k-means clustering is applied to the geochemical data matrix that automatically finds three hidden geothermal signatures representing modestly, moderately, and highly confident geothermal prospects. The algorithm also evaluated the probability of occurrence of these types of resources through the studied region. There is a consistency between regional geothermal prospectivity as estimated by our ML methodology and the traditional play fairway analysis conducted over a portion of the study area. We also identify the dominant data attributes associated with each signature. Finally, our ML analyses allow us to reconstruct attributes from sparse into continuous over the study domain. The predicted continuous attributes can be used for future detailed geothermal explorations in the Great Basin.

Three-way decision-based tri-training with entropy minimization

The three-way decision (TWD) theory is an effective methodology and philosophy for thinking in three and has been successfully applied to knowledge reasoning and decision making. However, limited research has been devoted to learning from partially labeled data with discrete and continuous attributes using TWD. In this study, we propose a TWD-based tri-training model for partially labeled data with heterogeneous attributes. First, a measure of semi-supervised neighborhood mutual information is defined, based on which a heuristic algorithm is developed to generate an optimal semi-supervised reduct of partially labeled data. Then, a tri-training model is trained on the original view along with two views transformed by data discretization and principal component analysis, and the strategy of TWD with entropy minimization is further introduced to classify unlabeled data into useful, uncertain, and useless samples, whereas the multiview tri-training model is iteratively retrained on only a certain number of useful samples with low entropy to improve the performance. Finally, the effectiveness of the proposed model is theoretically analyzed from the perspective of noise learning. The experimental results of semi-supervised attribute reduction and semi-supervised classification on UCI datasets show that our method is effective in handling partially labeled data and outperforms supervised models trained on all data with full supervision.

Intrusion Detection System Using K-Means and Edited Nearest Neighbour Algorithm

In imbalanced network traffic, malicious cyber-attacks can often hide in large amounts of normal data. It exhibits a high degree of stealth and obfuscation in cyberspace, making it difficult for Network Intrusion Detection System (NIDS) to ensure the accuracy and timeliness of detection. This paper researches machine learning and deep learning for intrusion detection in imbalanced network traffic. It proposes a novel Difficult Set Sampling Technique (DSSTE) algorithm to tackle the class imbalance problem. First, use the Edited Nearest Neighbor (ENN) algorithm to divide the imbalanced training set into the difficult set and the easy set. Next, use the K- Means algorithm to compress the majority samples in the difficult set to reduce the majority. Zoom in and out the minority samples’ continuous attributes in the difficult set synthesize new samples to increase the minority number. Finally, the easy set, the compressed set of majority in the difficult, and the minority in the difficult set are combined with its augmentation samples to make up a new training set. The algorithm reduces the imbalance of the original training set and provides targeted data augment for the minority class that needs to learn. It enables the classifier to learn the differences in the training stage better and improve classification performance. To verify the proposed method, we conduct experiments on the classic intrusion dataset NSL-KDD. We use classical classification models: random forest(RF), Support Vector Machine (SVM), XGBoost, Long and Short- term Memory (LSTM), Adaboost, AlexNet, Mini- VGGNet.

Belief Entropy Tree and Random Forest: Learning from Data with Continuous Attributes and Evidential Labels

As well-known machine learning methods, decision trees are widely applied in classification and recognition areas. In this paper, with the uncertainty of labels handled by belief functions, a new decision tree method based on belief entropy is proposed and then extended to random forest. With the Gaussian mixture model, this tree method is able to deal with continuous attribute values directly, without pretreatment of discretization. Specifically, the tree method adopts belief entropy, a kind of uncertainty measurement based on the basic belief assignment, as a new attribute selection tool. To improve the classification performance, we constructed a random forest based on the basic trees and discuss different prediction combination strategies. Some numerical experiments on UCI machine learning data set were conducted, which indicate the good classification accuracy of the proposed method in different situations, especially on data with huge uncertainty.

Classification Algorithm for Library Electronic Documents Based on Continuous Attribute

The automatic classification of document data will occupy an increasingly important position in digital libraries. Generally, the kernel method based on support vector machine is used to classify literature data on the standard test set, which has some shortcomings. In order to solve these problems, vocabulary expansion is used to preprocess the document vector to obtain a small but precise, orthogonal, and unambiguous new document vector; the document vector is sorted according to semantics to improve the access and calculation speed; the document is mapped to Lz with the help of wavelet kernel space for document classification. This paper analyzes the existing continuous attribute discretization methods in detail, discusses how to reduce the loss of information in the discretization process, and proposes a low-frequency discretization (LFD) algorithm based on the attribute low-frequency region. This method effectively reduces data loss by setting the segmentation point in the attribute interval with lower frequency, and through the research and analysis of the existing association rule mining algorithm, this paper combines low-frequency discretization, weighted multiple minimum support, and full confidence, and a weighted multiple minimum support association rule mining algorithm based on low-frequency discretization (WM-SamplingHT) is proposed. The algorithm first uses the low-frequency discretization algorithm to discretize the continuous attributes, then sets the respective weights and minimum support for the data items when mining frequent itemsets, removes the false patterns through the full confidence, and then obtains cleaner frequent itemsets. Using the real classification data of China Academic Journals Network, it is verified from the perspectives of abstract information and full-text documents. The results show that this method is superior to the nuclear method and has certain theoretical research and practical applications.

Rough-Set-Theory-Based Classification with Optimized k-Means Discretization

The discretization of continuous attributes in a dataset is an essential step before the Rough-Set-Theory (RST)-based classification process is applied. There are many methods for discretization, but not many of them have linked the RST instruments from the beginning of the discretization process. The objective of this research is to propose a method to improve the accuracy and reliability of the RST-based classifier model by involving RST instruments at the beginning of the discretization process. In the proposed method, a k-means-based discretization method optimized with a genetic algorithm (GA) was introduced. Four datasets taken from UCI were selected to test the performance of the proposed method. The evaluation of the proposed discretization technique for RST-based classification is performed by comparing it to other discretization methods, i.e., equal-frequency and entropy-based. The performance comparison among these methods is measured by the number of bins and rules generated and by its accuracy, precision, and recall. A Friedman test continued with post hoc analysis is also applied to measure the significance of the difference in performance. The experimental results indicate that, in general, the performance of the proposed discretization method is significantly better than the other compared methods.

Rule-Based Classification Based on Ant Colony Optimization: A Comprehensive Review

The Ant Colony Optimization (ACO) algorithms have been well-studied by the Operations Research community for solving combinatorial optimization problems. A handful of researchers in the Data Science community have successfully implemented various ACO methodologies for rule-based classification. This family of ACO algorithms is referred to as AntMiner algorithms. Due to the flexibility of the framework, and the availability of alternative strategies at the modular level, a systematic review on the AntMiner algorithms can benefit the broader community of researchers and practitioners interested in highly interpretable classification techniques. In this paper, we provided a comprehensive review of each module of the AntMiner algorithms. Our motivation is to provide insight into the current practices and future research scope in the context of the rule-based classification. Our discussions address ACO methodologies, rule construction strategies, candidate selection metrics, rule quality evaluation functions, rule pruning strategies, methods to address continuous attributes, parameter selection, and experimental settings. This review also reports a summary of real-life implementations of the rule-based classifiers in diverse domains including medical, genetics, portfolio analysis, geographic information system (GIS), human-machine interaction (HMI), autonomous driving, ICT, quality, and reliability engineering. These implementations demonstrate the potential application domains that can be benefitted from the methodological contributions to the rule-based classification technique.

Mixed-type data generation method based on generative adversarial networks

Data-driven based deep learing has become a key research direction in the field of artificial intelligence. Abundant training data is a guarantee for building efficient and accurate models. However, due to the privacy protection policy, research institutions are often limited to obtain a large number of training data, which would lead to a lack of training sets circumstance. In this paper, a mixed-type data generation model based on generative adversarial networks is proposed to synthesize fake data that have the same distribution with the real data, so as to supplement the real data and increase the number of available samples. The model first pre-trains the autoencoder which maps given dataset into a low-dimensional continuous space. Then, the generator constructed in the low-dimension space is obtained by training it adversarially with discriminator constructed in the original space. Since the constructed discriminator not only consider the loss of the continuous attributes but also the labeled attributes, the generator nets formed by the generator and the decoder can effectively learn the intrinsic distribution of the mixed data. We evaluate the proposed method both in the independent distribution of the attribute and in the relationship of the attributes, and the experiment results show that the proposed generate method has a better performance in preserve the intrinsic distribution compared with other generation algorithms based on deep learning.

Comparison of discretization methods for classifier decision trees and decision rules on medical data sets

Data sets in real life are given by real numbers in databases. On the other hand, many data mining methods like association rules and induction rules require only discrete attributes. For this reason, it is necessary to convert data sets with continuous attributes into data sets with discrete attributes. Discretization process is reducing the number of values for a given continuous attribute by dividing the range of the attribute into intervals. In this paper, eight discretization methods are presented with JRip, OneR, J48, and Part classifier algorithms of rules and tress. Experiments include a ten-fold cross validation provided on the basis of real-life data sets from the UCI repository. We show that discretization is important step to significantly increase the classification results of these algorithms. Finally, as a result of the study, it was seen that MDL and J48, CAIM and Jrip and Extended Chi and J48 methods gave the highest accuracy for PIMA, WBC and DERMA data sets, respectively.

A novel approach for discretizing continuous attributes based on tree ensemble and moment matching optimization

A novel approach for discretizing continuous attributes based on tree ensemble and moment matching optimization

GBNRS: A Novel Rough Set Algorithm for Fast Adaptive Attribute Reduction in Classification

Feature reduction is an important aspect of Big Data analytics on today&#x2019;s ever-larger datasets. Rough sets are a classical method widely applied in attribute reduction. Most rough set algorithms use the <i>priori</i> domain knowledge of a dataset to process continuous attributes through using a membership function. Neighborhood rough sets (NRS) replace the membership function with the concept of neighborhoods, allowing NRS to handle scenarios where no <i>a priori</i> knowledge is available. However, the neighborhood radius of each object in NRS is fixed, and the optimization of the radius depends on grid searching. This diminishes both the efficiency and effectiveness, leading to a time complexity of not lower than <inline-formula><tex-math notation="LaTeX">$O(N^2)$</tex-math><alternatives><mml:math><mml:mrow><mml:mi>O</mml:mi><mml:mo>(</mml:mo><mml:msup><mml:mi>N</mml:mi><mml:mn>2</mml:mn></mml:msup><mml:mo>)</mml:mo></mml:mrow></mml:math><inline-graphic xlink:href="xia-ieq1-2997039.gif"/></alternatives></inline-formula>. To resolve these limitations, granular ball neighborhood rough sets (GBNRS), a novel NRS method with time complexity <inline-formula><tex-math notation="LaTeX">$O(N)$</tex-math><alternatives><mml:math><mml:mrow><mml:mi>O</mml:mi><mml:mo>(</mml:mo><mml:mi>N</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math><inline-graphic xlink:href="xia-ieq2-2997039.gif"/></alternatives></inline-formula>, is proposed. GBNRS adaptively generates a different neighborhood for each object, resulting in greater generality and flexibility in comparison to standard NRS methods. GBNRS is compared with the current state-of-the-art NRS method, FARNeMF, and find that GBNRS obtains both higher performance and higher classification accuracy on public benchmark datasets. All code has been released in the open source GBNRS library at <uri>http://www.cquptshuyinxia.com/GBNRS.html</uri>.

Effect of Daily Multidisciplinary Team Reflection in Ambulatory Care: A Qualitative Analysis.

PurposeResearch suggests that daily previsit team huddling leads to improved teamwork, however there are no reports on the effect and mechanism of daily team reflection in primary care practice. To help healthcare professionals to integrate fragmented information and to enhance continuity, we previously developed and implemented a daily multidisciplinary team reflection (DMDTR) approach. The present study aimed to construct a conceptual diagram illustrating how participation in DMDTR influence medical care in an ambulatory clinic.Participants and MethodsIn this qualitative study, we analyzed data from semi-structured interviews (n = 12) conducted with physicians, nurses, and medical assistants who regularly participated in DMDTR. Modified grounded theory approach (M-GTA) was adopted as the analysis method, and structure-construction qualitative research method (SCQRM) was used as a meta-theory to determine the number of cases correlatively with the research objectives.ResultsOur conceptual diagram demonstrated that, through DMDTR, team members shared fragmented information to gain a more holistic view of the patient and discuss diverse points of view, a process that resulted in real-time problem solving through collaboration. This team approach showed benefit to each discipline and improves the care attributes of contextuality, coordination, accessibility, accountability, continuity, and comprehensiveness of care.ConclusionWe generated a conceptual diagram showing that DMDTR embodies a team approach that leads to improved care attributes, which results in enhanced quality of primary care.

Data stream classification with ant colony optimisation

Data stream mining has recently emerged in response to the rapidly increasing continuous data generation. While the majority of Ant Colony Optimisation (ACO) rule induction algorithms have proved to be successful in producing both accurate and comprehensive classification models in nonstreaming (batch) settings, currently ACO-based algorithms for classification problems are not suited to be applied to data stream mining. One of the main challenges is the iterative nature of ACO algorithms, where many procedures—for example, heuristic calculation, selection of continuous attributes, pruning—require multiple passes through the data to create a model. In this paper, we present a new ACO-based algorithm for data stream classification. The proposed algorithm, called Stream Ant-Miner (sAnt-Miner), uses a novel hybrid pheromone model combining both a traditional construction graph and solution archives models to efficiently handle a large number of mixed-type (nominal and continuous) attributes directly without the need for additional procedures, reducing the computational time required to complete an iteration of the algorithm. Our results show that sAnt-Miner produces statistically significant concise models compared with state-of-the-art rule induction data stream algorithms, without negative effects on their predictive accuracy.

Semi and Nonparametric Conditional Probability Density, a Case Study of Pedestrian Crashes

Background: Kernel-based methods have gained popularity as employed model residual’s distribution might not be defined by any classical parametric distribution. Kernel-based method has been extended to estimate conditional densities instead of conditional distributions when data incorporate both discrete and continuous attributes. The method often has been based on smoothing parameters to use optimal values for various attributes. Thus, in case of an explanatory variable being independent of the dependent variable, that attribute would be dropped in the nonparametric method by assigning a large smoothing parameter, giving them uniform distributions so their variances to the model’s variance would be minimal. Objectives: The objective of this study was to identify factors to the severity of pedestrian crashes based on an unbiased method. Especially, this study was conducted to evaluate the applicability of kernel-based techniques of semi- and nonparametric methods on the crash dataset by means of confusion techniques. Methods: In this study, two non- and semi-parametric kernel-based methods were implemented to model the severity of pedestrian crashes. The estimation of the semi-parametric densities is based on the adoptive local smoothing and maximization of the quasi-likelihood function, which is similar somehow to the likelihood of the binary logit model. On the other hand, the nonparametric method is based on the selection of optimal smoothing parameters in estimation of the conditional probability density function to minimize mean integrated squared error (MISE). The performances of those models are evaluated by their prediction power. To have a benchmark for comparison, the standard logistic regression was also employed. Although those methods have been employed in other fields, this is one of the earliest studies that employed those techniques in the context of traffic safety. Results: The results highlighted that the nonparametric kernel-based method outperforms the semi-parametric (single-index model) and the standard logit model based on the confusion matrices. To have a vision about the bandwidth selection method for removal of the irrelevant attributes in nonparametric approach, we added some noisy predictors to the models and a comparison was made. Extensive discussion has been made in the content of this study regarding the methodological approach of the models. Conclusion: To summarize, alcohol and drug involvement, driving on non-level grade, and bad lighting conditions are some of the factors that increase the likelihood of pedestrian crash severity. This is one of the earliest studies that implemented the methods in the context of transportation problems. The nonparametric method is especially recommended to be used in the field of traffic safety when there are uncertainties regarding the importance of predictors as the technique would automatically drop unimportant predictors.