Towards Application of One-Class Classification Methods to Medical Data

Abstract

In the problem of one-class classification (OCC) one of the classes, the target class, has to be distinguished from all other possible objects, considered as nontargets. In many biomedical problems this situation arises, for example, in diagnosis, image based tumor recognition or analysis of electrocardiogram data. In this paper an approach to OCC based on a typicality test is experimentally compared with reference state-of-the-art OCC techniques—Gaussian, mixture of Gaussians, naive Parzen, Parzen, and support vector data description—using biomedical data sets. We evaluate the ability of the procedures using twelve experimental data sets with not necessarily continuous data. As there are few benchmark data sets for one-class classification, all data sets considered in the evaluation have multiple classes. Each class in turn is considered as the target class and the units in the other classes are considered as new units to be classified. The results of the comparison show the good performance of the typicality approach, which is available for high dimensional data; it is worth mentioning that it can be used for any kind of data (continuous, discrete, or nominal), whereas state-of-the-art approaches application is not straightforward when nominal variables are present.

Toward Detecting Infection Incidence in People With Type 1 Diabetes Using Self-Recorded Data (Part 1): A Novel Framework for a Personalized Digital Infectious Disease Detection System

“…In a similar fashion, novelty detection–based algorithms can be other alternatives for detecting novel microevents relying on either supervised, semisupervised, and unsupervised approaches [ 4 , 66 , 67 , 81 ]. Different categories of novelty detection approaches could be exploited for detecting infection-induced deviations in blood glucose dynamics, including approaches based on statistical techniques [ 68 ], prediction, density [ 82 - 85 ], distance [ 67 , 86 ], classification or domain [ 4 , 62 , 63 , 87 - 92 ], clustering [ 62 , 93 ], and ensemble [ 67 , 68 , 80 - 82 , 85 , 86 , 92 - 95 ] .…”

HIDROID: Prototyping a Behavioral Host-Based Intrusion Detection and Prevention System for Android

“…1) we develop HIDROID as an IDPS application for Android, describing its full functional and implementation details; 2) we complement the detection engine of the proposed Intrusion Detection System (IDS), in [10], with an Intrusion Prevention System (IPS), paving the way towards a complete IDPS solution; 3) we evaluate the performance of HIDROID while running on a real-life mobile device and present our experimental results. In order to keep the computation burden of HIDROID on a mobile device at minimum, without loss of generality, we implement a couple of lightweight Machine Learning (ML) and statistical algorithms, namely one-class K-means [11] -a variant of K-means clustering algorithm with only one cluster encompassing the benign data -and the univariate Gaussian algorithm for anomaly detection [12], [10]. The results published in our previous work, in [10], were based on simulations only, while this work presents experimental results from our implemented prototype.…”

A Novel Approach for Continuous Health Status Monitoring and Automatic Detection of Infection Incidences in People With Type 1 Diabetes Using Machine Learning Algorithms (Part 2): A Personalized Digital Infectious Disease Detection Mechanism

“…In this case, semisupervised anomaly detection, that is, one-class classification, is preferred given that it only requires characterizing what is believed to be normal (target data instances) to detect the abnormal (nontarget data instances) [ 7 ]. Under certain circumstances, for example, medical domain, obtaining and demarcating the anomalous (nontarget) data instances can become very difficult, expensive, and time consuming, if not impossible [ 7 , 13 ]. For instance, assume a health diagnostic and monitoring system that detects health changes in an individual by tracking the individual’s physiological parameters, where the current health status is examined based on a set of parameters, and raises a notification alarm when the individual health deteriorates [ 12 ].…”

“…This is because demarcating the exact boundaries between normal and abnormal health conditions is very challenging given that each pathogen has a different effect on the individual physiology. The one-class classifiers–based anomaly detection methods can be roughly grouped into 3 main groups: boundary and domain-based, density-based, and reconstruction-based methods based on how their internal function is defined and the approach used for minimization [ 3 , 10 , 12 , 13 , 15 , 16 ]. These models take into account different characteristics of the data set, and depending on the data set under consideration, these models could achieve different generalization performance, overfitting, and bias [ 12 ].…”

“…Both semisupervised and unsupervised methods have been used in various applications to detect anomalous instances [ 1 , 7 , 10 , 16 ]. In particular, these methods have been popular in the medical domain owing to their suitability for such applications, where there is lack of a sufficient data set for the other classes [ 13 ]. Accordingly, considering the difficulty and expense of obtaining enough sample data sets for the infection days from people with type 1 diabetes, a one-class classifier and unsupervised models are proposed for detecting infection incidence in people with type 1 diabetes.…”