Search

Serum free fatty acids (FFA) are generally elevated in obesity. The gut microbiota is involved in the host energy metabolism through the regulation of body fat storage, and a link between diet, FFA and the intestinal microbiota seems to exist. Our aim was to explore the interaction among serum FFA levels, gut microbiota, diet and obesity through a model regression tree in 66 subjects (age 52.7 ± 11.2 y) classified according to Body Mass Index (BMI). Total and individual FFA were analyzed by colorimetric enzymatic assay and methyl-tert-butylether-based extraction protocol (MTBE), respectively. Microbiota was determined by qPCR and diet through a food frequency questionnaire. Statistical analyses were performed, and predictive factors for obesity were obtained via classification by decision trees using machine learning methods. An obese-linked FFA profile was characterized by decreased eicosapentaenoic (EPA) and increased linoleic, gamma-linolenic and palmitic acids levels simultaneously. Serum EPA and gender were identified as the most significant variables with 100% and 80% of importance, respectively. Palmitic acid, Bifidobacterium and Faecalibacterium explained >30%, followed by Bacteroides group with 20% and docosahexaenoic acid (DHA) almost with 15% of importance. Also, the regression tree model obtained for predicting obesity, showed a non-obese-linked profile, independently of gender, with serum EPA > 0.235 μg/mL and Bacteroides > 9.055 log n° cells per g of feces. Moreover, Faecalibacterium and Bifidobacterium seemed to play an important role by complementing the levels of FFA in predicting obesity in males and females, respectively., (Copyright © 2018 Elsevier Ltd. All rights reserved.)

Introducing intelligent safety devices in the roads would lead to enhance both the time of reaction and the traffic safety. Nevertheless, these intelligent devices are expensive, so choosing their location should be done carefully. This research is focused a decision support system to decide the placement of a specific safety device designed in a research project. This approach includes a feature selection stage, a model learning stage and the deployment stage. Decision models learn from real datasets with information related with accidents, classifying the samples as Fatal, Severe or Slight injury. Also, a case based risk index is proposed, so samples within the same label can be sorted. Therefore, in the deployment stage, each possible location is ranked and the user gets a feedback of the suitability of each of them to be considered for placing the intelligent safety device. The experimentation shows the proposal is valid provided the dataset for training includes enough granularity. However, it is shown that specific risk index should be designed for each road type and fork. [ABSTRACT FROM AUTHOR]

University dropout is a growing problem with considerable academic, social and economic consequences. Conclusions and limitations of previous studies highlight the difficulty of analyzing the phenomenon from a broad perspective and with bigger data sets. This paper proposes a new, machine-learning based method, able to examine the problem using a holistic approach. Advantages of this method include the lack of strong distribution hypothesis, the capacity for handling bigger data sets and the interpretability of the results. Results are consistent with previous research, showing the influence of personal and contextual variables and the importance of academic performance in the first year, but other factors are also highlighted with this model, such as the importance of dedication (part or full time), and the vulnerability of the students with respect to their age. Additionally, a comprehensive graphic output is included to make it easier to interpret the discovered rules. [ABSTRACT FROM AUTHOR]

Classification is a long standing problem in computing. There are two broad kinds of classifiers, frequency based and geometry based. Frequency based classifiers often ignore the geometry underlying the data. Conversely, geometry based classifiers take into account the frequency only indirectly. This paper presents a classification algorithm which considers explicitly geometric and statistical characteristics of the data and combines them into a class representation. Reported here are initial experiments with this algorithm using two well known data sets, both with and without noise. The results show that the proposed algorithm is less sensitive to the training data set than other classifiers. [ABSTRACT FROM AUTHOR]

This work develops a decision-supported system based on machine learning and scoring measures to discover the kind of female urinary incontinence (FUI) of a given patient. This system has two main branches. Each patient is characterized by a set of features (age, weight, number of childbirths, etc.). The first task consists of selecting the feature set which best defines each FUI class. This feature set is computed according to a some scoring measures. The patients characterized by the optimum feature set are then classified according to a Support Vector Machine classifier. The results are evaluated in terms of macroaccuracy, i.e. the mean of the percentages of correctly classified stress, mixed and sensory urge incontinence. [ABSTRACT FROM AUTHOR]

Abstract: This work develops a decision support system based on machine learning and scoring measures to determine the type of urinary incontinence in women with low urinary tract symptoms. This system has two main branches. The former consists of selecting the feature set which best defines the UI type from the set of features (age, weight, etc.) characterizing a patient. This feature set is computed from several scoring measures. The patients characterized by the optimum feature set are then classified according to C4.5 and SVM classifiers. The results are evaluated according to Sensitivity and Specificity evaluation measures. The management of the final system is simple and its performance is high, getting Sensitivities over 80% and Specificities near 100% for some configurations. [Copyright &y& Elsevier]

Text Categorization, which consists of automatically assigning documents to a set of categories, usually involves the management of a huge number of features. Most of them are irrelevant and others introduce noise which could mislead the classifiers. Thus, feature reduction is often performed in order to increase the efficiency and effectiveness of the classification. In this paper, we propose to select relevant features by means of a family of linear filtering measures which are simpler than the usual measures applied for this purpose. We carry out experiments over two different corpora and find that the proposed measures perform better than the existing ones. [ABSTRACT FROM AUTHOR]