A Variable Ranking Method for Machine Learning Models with Correlated Features: In-Silico Validation and Application for Diabetes Prediction

When building a predictive model for predicting a clinical outcome using machine learning techniques, the model developers are often interested in ranking the features according to their predictive ability. A commonly used approach to obtain a robust variable ranking is to apply recursive feature elimination (RFE) on multiple resamplings of the training set and then to aggregate the ranking results using the Borda count method. However, the presence of highly correlated features in the training set can deteriorate the ranking performance. In this work, we propose a variant of the method based on RFE and Borda count that takes into account the correlation between variables during the ranking procedure in order to improve the ranking performance in the presence of highly correlated features. The proposed algorithm is tested on simulated datasets in which the true variable importance is known and compared to the standard RFE-Borda count method. According to the root mean square error between the estimated rank and the true (i.e., simulated) feature importance, the proposed algorithm overcomes the standard RFE-Borda count method. Finally, the proposed algorithm is applied to a case study related to the development of a predictive model of type 2 diabetes onset.

Implementing Fuzzy TOPSIS on Project Risk Variable Ranking

Managing construction risks with a large number of risks with small impact can increase the additional effort and cost of inefficient construction. Therefore the variables need to be eliminated. The aim of this study is ranking the risk variable based on its frequency of occurrence by integrating time, cost, and quality criteria simultaneously and selecting the top ten variables with the order of the most significant impact. The risk variable ranking based on triple project objective of cost, time, and quality simultaneously is a challenge for particular projects or regions contributing to the risk context. A number of 127 qualitative risk variables of 14 factors occurring in a project to be eliminated require a method/technique. A fuzzy TOPSIS method involving linguistics data is proposed to capture vague conditions. Results show that the top ten rankings of risk variables based on integrating the different weights of cost, time, and quality are successfully identified by concluding that the labour factor is the most dominant variable affecting project risk in context the rehabilitation and reconstruction posttsunami disaster, especially in Aceh-Indonesia. The variables are lack of labour, unskilled labour, undisciplined labour, and low productivity of labour. This condition can differ from different risk contexts. This research is different from other studies that only review cost, time, and quality separately. We stated that to integrate all three criteria of cost, time, and quality simultaneously is more logic to analyze risk variable ranking.

Filter Variable Selection Algorithm Using Risk Ratios for Dimensionality Reduction of Healthcare Data for Classification

This research developed and tested a filter algorithm that serves to reduce the feature space in healthcare datasets. The algorithm binarizes the dataset, and then separately evaluates the risk ratio of each predictor with the response, and outputs ratios that represent the association between a predictor and the class attribute. The value of the association translates to the importance rank of the corresponding predictor in determining the outcome. Using Random Forest and Logistic regression classification, the performance of the developed algorithm was compared against the regsubsets and varImp functions, which are unsupervised methods of variable selection. Equally, the proposed algorithm was compared with the supervised Fisher score and Pearson’s correlation feature selection methods. Different datasets were used for the experiment, and, in the majority of the cases, the predictors selected by the new algorithm outperformed those selected by the existing algorithms. The proposed filter algorithm is therefore a reliable alternative for variable ranking in data mining classification tasks with a dichotomous response.

Deriving sonority from the structure, not the other way round: A Strict CV approach to consonant clusters

This paper aims to show that sonority-based generalizations on consonant phonotactics should directly follow from representations, not from stipulations on representations such as the commonly accepted licensing or government statements. The basic reason for this is that the second approach is both arbitrary and circular, as it entails a variable ranking of alleged well-formedness principles, if we want to explain, for example, why TR clusters are either tautosyllabic or heterosyllabic depending on the language. I argue instead for a representational alternative assuming that (i) consonants and vowels are universally segregated, and (ii) involve two parallel CVCV sequences – one on the C-plane, the other on the V-plane – (iii) which may differ in length. It is shown how the major sonority categories, and thereby the phonotactic constraints based on these categories, naturally result from how the two CVCV sequences are synchronized if the one on the C-plane is longer than the one on the V-plane. It will also be seen how the proposed structures naturally account for several processes such as liquid metathesis and deletion, vowel epenthesis, plosive fricativization, etc., while providing a means for measuring the relative likelihood of some of them on the basis of representational markedness.