RANDOM FORESTS SUPPLEMENTED BY PARALLEL FENCES PLOTS: A STUDY IN SPIROMETRIC DIAGNOSTICS
Journal Title: Colloquium Biometricum; Colloquium Biometricum (Online) - Year 2018, Vol 48, Issue
Abstract
We try to assess, what truly can be obtained when using only the spirometric measurements for the diagnosis of pulmonary disorders. Using spirometric data recorded for 202 patients and the Random Forests methodology, we obtained an algorithm permitting for a 100% correct classification of the learning sample, and for a 94% correct classification of the 125 'not-healthy' patients included into the validation sample. We propose, as a routine continuation, to verify the results - especially those for misclassified patients - by an independent multivariate graphical procedure called by us the parallel fences plot. By inspecting the patient's data vectors displayed in parallel fences plots it becomes evident if they are typical or not-typical for the 'healthy' or 'not-healthy' state of the patient. The conclusion is, that the spirogram can be used for a relatively safe diagnosis of majority of patients. Only those that do not pass the parallel fences plot test should undergo further pulmonary scrutinizing test.
Authors and Affiliations
Anna Bartkowiak, Jerzy Liebhart
Keywords
Related Articles
- EP ID EP535741
- DOI -
- Views 76
- Downloads 0
What story does the Durov diagram tell?
The Durov diagram is a popular visualization technique applied in hydrogeology to display the major ions as percentages of milli-equivalents in two ternary (trilinear) graphs that form an additional two-dimensional proje...
A REMARK ON ESTIMATING DEFECTIVENESS IN SAMPLING ACCEPTANCE INSPECTION
Summary The problem of estimating a probability of success in a Binomial model is considered. The classical estimator is compared with the estimator which uses the information about non-homogeneity of the sample. An appl...
D-OPTIMAL DESIGNS WITH NEGATIVE CORRELATED ERRORS BASED ON TERNARY DESIGNS: CONSTRUCTION
Here, we study new construction methods of the regular D-optimal chemical balance weighing design with equally negative correlated errors. The construction method is based on the set of the incidence matrices of the tern...
NEW CONSTRUCTION OF D-OPTIMAL WEIGHING DESIGNS WITH NON-NEGATIVE CORRELATIONS OF ERRORS
In this paper, the regular D-optimal chemical balance weighing designs with equally non-negative correlated errors are considered. Here we study the issues regard to the existence conditions of optimal designs. Presented...
APPLICATION OF MULTIVARIATE STATISTICAL ANALYZES FOR COMPARISON OF BETULA POLLEN SEASONS 2001-2016 IN LUBLIN
The aim of the study was finding similar groups of birch pollen season in Lublin in 2001-2016 based on the pollen seasons parameters such as start, end, duration, peak date, peak value and annual pollen sum. Cluster anal...