Efficient discovery of risk patterns in medical data

Delays in the completion of pvd production can be caused by several factors. Including the actual experience in the production of the difficulty of each process and color type, even the difficulty of the product type can also be affected. In this study, the prediction of the delay in the completion of pvd production was carried out using the C4.5 decision tree and Multilayer Perceptron data mining method approach using Production Results data at PT. Surya Toto Indonesia, whose results are expected to provide information and input for the company in making production plans in the future. The data testing method was carried out with 5 (five) testing times with different amounts of data to determine the level of consistency of accuracy obtained. C4.5 gives the results of a decision tree where the root is the color type and as the leaf is the product category, type type and order period. The average value of accuracy generated in the C4.5 decision tree method is 87.15%. While the Multilayer Perceptron obtained an average accuracy of 87.98%, which is greater than the decision tree method with a difference of 0.83%.

In this study, the decision tree (DT) technique is developed to assess the basketball strategies' training impact on physical education (PE). The recommended DT method effectively guesses the tactics of two teams based on the locations of basketball players with only a small amount of training data. Next, using data concerning ball possession, the interaction of these teams' tactics, and the distinctive features of basketball strategies, the DT approach changes the team's strategies.Therefore, effective findings from prediction could be produced because the suggested DT approach predicts the team strategies that meet these requirements. The suggested DT approach is implemented using the Python platform to analyze efficiency in terms of several metrics. Additionally, the DT method's effectiveness is contrasted with that of other approaches. Based on the results of the experiments, the suggested approach outperforms the current approaches in assessing the impact of basketball strategy instruction in physical education.

For situations where the data are drawn from reasonably homogeneous populations, traditional methods such as multiple regression typically yield insightful analyses. For situations where the data are drawn from more heterogeneous populations, decision tree approaches, such as Classification and Regression Trees (CART) and Generalized, Unbiased, Interaction, Detection, and Estimation (GUIDE), are more likely to recognize idiosyncratic subpopulations and interactions automatically. In contrast to CART, however, GUIDE yields models with better predictive performance for each subpopulation. This article extends the idea of GUIDE to handle analysis of covariance-type problems. This article compares GUIDE modeling to various decision tree methods and to multiple regression. The article identifies and discusses the relative advantages and disadvantages of multiple regression, CART, and GUIDE. GUIDE produces quality or reliability models that exhibit greater predictive accuracy than multiple regression or CART for complex, highly diverse populations. Also, GUIDE is readily applicable to many other areas, such as repairability and maintainability settings involving both qualitative and quantitative variables. A small case study of an engineered wood product, medium-density fiberboard, is presented to illustrate the application of GUIDE. Accepted in 2005 for a special issue on Reliability co-edited by Hoang Pham, Rutgers University; Dong Ho Park, Hallym University, Korea; and Richard Cassady, University of Arkansas.

The rapid adoption of Internet of Things (IoT) technologies in healthcare has given rise to the Internet of Medical Things (IoMT), which has transformed patient care and medical services. The IoMT, when combined with Fog Computing, provides a powerful paradigm for processing and analyzing healthcare data at the network edge. This paper proposes an innovative intelligent load balancing framework designed specifically for fog-enabled IoMT environments for optimizing resource utilization, improving system performance, and ensuring timely and efficient healthcare service delivery. The framework dynamically distributes computing tasks among fog nodes based on real-time parameters such as node capacity, latency, and workload. By combining machine learning (ML) models and data analytics, the system adapts to changing patterns in medical data, ensuring adaptive load distribution and faster response times. The proposed framework addresses the unique challenges facing healthcare applications, such as low latency and energy consumption in data transmission.

Heart diseas remains major cause of death in world wide and early diagnosis is vital to effective clinical treatment. Models that are capable of capturing complex patterns in medical data are needed in Heart disease prediction. This paper presents a hybrid quantum-classical setup that consists of a Quantum Neural Network (QNN), Quantum SVM kernel modeling and classical learners stacked framework. In the experiment with the cleaned UCI heart disease dataset and $\mathbf{1 0}$-fold cross-validation, the proposed model has an accuracy of 91.48, F1-score of 0.8955 and AUC of 0.9395, surpassing eight classical baselines. SHAP and LIME are also interpretable, which emphasizes the most influential features. A low scale of samples is also run on IBM Quantum hardware where mitigation is performed by resiliency level to ensure pragmatic feasible outcome. The findings show that quantum components can improve nonlinear feature representation, which proves the possibilities of hybrid quantumbased medical diagnosis.

PurposeThe emerging field of educational data mining provides an opportunity to process large-scale data emerging from higher education institutions (HEIs) into reliable knowledge. The purpose of this paper is to examine factors influencing persistence amongst students enrolled in a Chemistry major at a South African university using enrolment data.Design/methodology/approachThe sample consisted of 511 students registered for a Chemistry major beginning in 2012, 2013 and 2014. Descriptive statistics in counts and percentages and classification (decision) tree methods were used in the analysis.FindingsGraduation from the Chemistry major is likely to occur after 4 years, which is regulation time plus 1 year, whilst departure mainly occurs in the first year of study. Classification tree modelling demonstrated that first year accumulated credits (FYAC), gender, financial aid status and school quintile were the factors associated with persistence. FYAC was the most critical factor.Research limitations/implicationsAlthough this study has many strengths, significantly the use of data mining methods to classify students, some limitations might affect how the results are interpreted. First, the analysis focused on a one-degree major in one institution, which leads to the suspicion that the observed results are discipline or institution-specific. Thus, the findings cannot be generalised to other contexts or disciplines. Second, with so many potential factors influencing student persistence, the analysis presented in this paper, which was limited to the covariates obtained in the institutional dataset the authors used, is by no means exhaustive. There is the possibility that some factors, which are not included in the present analysis, might have more predictive power.Originality/valueGlobally, university administrators are interested in predicting student outcomes and understanding the intricate balance between enrolment and throughput. Thus, whilst the findings from this study have an institutional focus, they resonate with other HEIs and present an alternative and highly visual way of identifying specific combinations of factors associated with persistence. The results from a classification tree model can also classify students at risk and inform the development of interventions that will support them.

In the Langcang (Upper-Mekong) watershed, degraded watershed ecosystems in upland areas threaten cultivation practices, water resources, and dam development downstream. Assessment of ecosystem services and the factors that threaten them is an important first step to support watershed management. This, however, requires detailed information that is often missing in mountainous regions. To overcome this, in this paper, we adopt a decision tree approach to assess protection, biodiversity, and production services in two mountainous watersheds (Fengqing and Xiaojie) of the Lancang River Basin. Decision tree rules were built on the basis of field surveys, available references, ecosystem maps derived from remote sensing, expert knowledge, basic topographic information, and community interviews. Decision tree results showed that forest cover and agro-forestry practices contribute greatly to improved ecosystem functioning in the Fengqing Catchment compared to the Xiaojie Catchment. The results were consistent with field observations. The decision tree method proved to be a suitable and flexible tool for the rapid assessment of watershed ecosystem services, for highlighting those areas that need more in-depth research, and for guiding watershed and ecosystem management.

Computational approach to predict peptide binding to major histocompatibility complex (MHC) is crucial for vaccine design since these peptides can act as a T-Cell epitope to trigger immune response. There are two main branches for peptide prediction methods; structural and data mining approaches. These methods can be successfully used for prediction of T-Cell epitopes in cancer, allergy and infectious diseases. In this paper, association rule mining methods are implemented to generate rules of peptide selection by MHCs. To capture the binding characteristics, modified rule mining and data transformation methods are implemented in this paper. Peptides are known to bind to the same MHC show sequence variability, to capture this characteristic, we used a reduced amino acid alphabet by clustering amino acids according to their physico-chemical properties. Using the classification of amino acids and the OR-operator to combine the rules to reflect that different amino acid types and positions along the peptide may be responsible for binding are the innovations of the method presented. We can predict MHC Class-I binding with 75-97% coverage and 76-100% accuracy.

Decision trees have been successfully used for years in many medical decision making applications. Transparent representation of acquired knowledge and fast algorithms made decision trees one of the most often used symbolic machine learning approaches. This paper concentrates on the problem of separating acute appendicitis, which is a special problem of acute abdominal pain, from other diseases that cause acute abdominal pain by use of an decision tree approach. Early and accurate diagnosing of acute appendicitis is still a difficult and challenging problem in everyday clinical routine. An important factor in the error rate is poor discrimination between acute appendicitis and other diseases that cause acute abdominal pain. This error rate is still high, despite considerable improvements in history-taking and clinical examination, computer-aided decision-support, and special investigation such as ultrasound. We investigated three databases of different size with cases of acute abdominal pain to complete this task as successful as possible. The results show that the size of the database does not necessary directly influence the success of the decision tree built on it. Surprisingly we got the best results from the decision trees built on the smallest and the biggest database, where the database with medium size (relative to the other two) was not so successful. Despite this we were able to produce decision tree classifiers that were capable of producing correct decisions on test data sets with accuracy up to 84%, sensitivity to acute appendicitis up to 90%, and specificity up to 80% on the same test set.

Optimal maintenance-policies for deteriorating systems under various maintenance strategies : C. Teresa Lam and R. H. Yeh. IEEE Transactions on Reliability, 43, 3, 423 (September 1994)

Cervical cancer is a disease that develops in the cervix’s tissue. Cervical cancer mortality is being reduced due to the growth of screening programmers. Cervical cancer screening is a big issue because the majority of cervical cancer screening treatments are invasive. Hence, there is apprehension about standard screening procedures, as well as the time it takes to learn the results. There are different methods for detecting problems in the cervix using Pap (Papanicolaou-stained) test, colposcopy, Computed Tomography (CT), Magnetic Resonance Image (MRI) and ultrasound. To obtain a clear sketch of the infected regions, using a decision tree approach, the captured image has to be segmented and analyzed. The goal of creating a decision tree is to establish prediction model that anticipate the feature vector based on the input variable. This paper deals with investigating various techniques of segmentation for detecting the cervical cancer. It proposes a novel method to develop an assistance system for the detection diagnosis of cervical cancer, based on work of Martin, Byriel and Norup. The analysis is focused on Pap smear pictures of single cells. Smear testing is a method of detecting abnormalities in the blood. Image processing is an effective method for extracting data. It is used to determine the size of cervical carcinoma and the length of the uterus. Martin’s database, which is open source and utilised for analysis and validation, is obtainable for research purposes. Cervical malignancy information utilizing three grouping strategies to anticipate the disease and afterward analyzed the outcomes showed that choice tree is the best classifier indicator with the test dataset. Further investigations ought to be led to improve execution.

Automatic Learning Approaches for On-Line Transient Stability Preventive Control of the Hydro-Quebec System: Part I. Decision tree approaches

A method for mining association rules that reflect the behaviors of past users is proposed for an adaptive search engine. The logs of the users' retrieving behaviors are described with the resource description framework model, from which association rules that reflect successful retrieving behaviors are extracted. The extracted rules are used to improve the performance of a metadata-based search engine. The document repository with adaptive hybrid search engine is also developed based on the proposed method. The repository consists of a document registration module, hybrid search engine, and reasoning base. The document registration module is designed to reduce the cost of adding metadata to documents, and the hybrid search engine combines full-text search with metadata-based search engine to improve the recall of retrieval result. The reasoning base is implemented based on the association rule mining method, which contributes to improve both precision and recall of the hybrid search engine. Experiments are performed with a virtual user model, of which results show that appropriate rules can be extracted with the proposed method. The proposed technologies will contribute to realize the concept of humatronics in terms of establishing symmetric relation between humans and systems, as well as sharing information, knowledge, and experiences via computer networks

In this paper we develop a new approach for learning decision trees and multivariate polynomials via interpolation of multivariate polynomials. This new approach yields simple learning algorithms for multivariate polynomials and decision trees over finite fields under any constant bounded product distribution. The output hypothesis is a (single) multivariate polynomial that is an /spl epsiv/-approximation of the target under any constant bounded product distribution. The new approach demonstrates the learnability of many classes under any constant bounded product distribution and using membership queries, such as j-disjoint DNF and multivariate polynomial with bounded degree over any field. The technique shows how to interpolate multivariate polynomials with bounded term size from membership queries only. This in particular gives a learning algorithm for O(log n)-depth decision tree from membership queries only and a new learning algorithm of any multivariate polynomial over sufficiently large fields from membership queries only. We show that our results for learning from membership queries only are the best possible.

Decision tree classification is one of the most efficient methods for obtaining land use/land cover (LULC) information from remotely sensed imageries. However, traditional decision tree classification methods cannot effectively eliminate the influence of mixed pixels. This study aimed to integrate pixel unmixing and decision tree to improve LULC classification by removing mixed pixel influence. The abundance and minimum noise fraction (MNF) results that were obtained from mixed pixel decomposition were added to decision tree multi-features using a three-dimensional (3D) Terrain model, which was created using an image fusion digital elevation model (DEM), to select training samples (ROIs), and improve ROI separability. A Landsat-8 OLI image of the Yunlong Reservoir Basin in Kunming was used to test this proposed method. Study results showed that the Kappa coefficient and the overall accuracy of integrated pixel unmixing and decision tree method increased by 0.093% and 10%, respectively, as compared with the original decision tree method. This proposed method could effectively eliminate the influence of mixed pixels and improve the accuracy in complex LULC classifications.