The Balanced Accuracy and Its Posterior Distribution

Brodersen, Kay H.; Ong, Cheng Soon; Stephan, Klaas Ε.; Buhmann, Joachim M.

doi:10.1109/icpr.2010.764

Cited by 1,061 publications

(641 citation statements)

References 5 publications

Supporting

Mentioning

631

Contrasting

Unclassified

Order By: Relevance

“…Classification results are summarized as balanced recognition accuracy (i.e. average of sensitivity and specificity) [49]. Additionally, in order to explore the relative importance of all features in the classification problem, we employed a support vector machine recursive feature elimination (SVM-RFE) procedure in a wrapper approach (RFE was performed on the training set of each fold and we computed the median rank for each feature over all folds).…”

Section: Single-subject Classification and Feature Selectionmentioning

confidence: 99%

Assessment of spontaneous cardiovascular oscillations in Parkinson's disease

Valenza

Orsolini

Diciotti

et al. 2016

Biomedical Signal Processing and Control

View full text Add to dashboard Cite

Parki so 's disease PD has ee reported to i ol e postga glio i s patheti failure and a wide spectrum of autonomic dysfunctions including cadiovascular, sexual, bladder, gastrointestinal and sudo-motor abnormalities. While these symptoms may have a significant impact on daily activities, as well as quality of life, the evaluation of autonomic nervous system (ANS) dysfunctions relies on a large and expensive battery of autonomic tests only accessible in highly specialized laboratories. In this paper we aim to devise a comprehensive computational assessment of disease-related heartbeat dynamics based on instantaneous, time-varying estimates of spontaneous (resting state) cardiovascular oscillations in PD. To this end, we combine standard ANS-related heart rate variability (HRV) metrics with measures of instantaneous complexity (dominant Lyapunov exponent and entropy) and higher-order statistics (bispectra). Such measures are computed over 600-s recordings acquired at rest in 29 healthy subjects and 30 PD patients. The only significant group-wise differences were found in the variability of the dominant Lyapunov exponent. Also, the best PD vs. healthy controls classification performance (balanced accuracy: 73.47%) was achieved only when retaining the time-varying, non-stationary structure of the dynamical features, whereas classification performance dropped significantly (balanced accuracy: 61.91%) when excluding variability-related features. Additionally, both linear and nonlinear model features correlated with both clinical and neuropsychological assessments of the considered patient population.Our results demonstrate the added value and potential of instantaneous measures of heartbeat dynamics and its variability in characterizing PD-related disabilities in motor and cognitive domains.3

show abstract

Section: Single-subject Classification and Feature Selectionmentioning

confidence: 99%

Assessment of spontaneous cardiovascular oscillations in Parkinson's disease

Valenza

Orsolini

Diciotti

et al. 2016

Biomedical Signal Processing and Control

View full text Add to dashboard Cite

show abstract

“…For criticality assessment with respect to the complete traffic scene, the individual elements need to be combined. Considering all f predicted vehicles and the static environment, the probability of the combined event C k (T P ): The ego vehicle collides with at least one vehicle or the static driving environment at least once within the prediction horizon {k:k +T P }, evaluated at time step k, can be derived as 46) in which the first step follows from de Morgan's laws and the last step from the stochastic independence of individual (not mutually exclusive) collision events. This independence is a consequence of the independence assumption of all vehicle trajectories, so that the probability of the ego vehicle colliding with one object is not influenced by the probability of colliding with another.…”

Section: Combinationmentioning

confidence: 99%

Bayesian environment representation, prediction, and criticality assessment for driver assistance systems

Schreier

2017

At - Automatisierungstechnik

View full text Add to dashboard Cite

The dissertation approaches the questions i) how to represent the driving environment in an environment model, ii) how to obtain such a representation, and iii) how to predict the traffic scene for criticality assessment. Bayesian inference provides the common framework of all designed methods. First, Parametric Free Space (PFS) maps are introduced, which compactly represent the vehicle environment in form of relevant, drivable free space. They are obtained by a novel method for grid mapping and tracking in dynamic environments. In addition, a maneuver-based, long-term trajectory prediction and criticality assessment system is introduced and the application of all methods within the advanced driver assistance system PRORETA 3 is described.

show abstract

“…Balanced accuracy (BACC) : The balanced accuracy, which can be defined as the average accuracy obtained on either class [25].…”

Section: %mentioning

confidence: 99%

Skin Lesion Classification using Machine Learning Algorithms

Özkan

Köklü

2017

ijisae

View full text Add to dashboard Cite

Melanoma is a deadly skin cancer that breaks out in the skin's pigment cells on the skin surface. Melanoma causes 75% of the skin cancer-related deaths. This disease can be diagnosed by a dermatology specialist through the interpretation of the dermoscopy images in accordance with ABCD rule. Even if dermatology experts use dermatological images for diagnosis, the rate of the correct diagnosis of experts is estimated to be 75-84%. The purpose of this study is to pre-classify the skin lesions in three groups as normal, abnormal and melanoma by machine learning methods and to develop a decision support system that should make the decision easier for a doctor. The objective of this study is skin lesions based on dermoscopic images PH 2 datasets using 4 different machine learning methods namely; ANN, SVM, KNN and Decision Tree. Correctly classified instances were found as 92.50%, 89.50%, 82.00% and 90.00% for ANN, SVM, KNN and DT respectively. The findings show that the system developed in this study has the feature of a medical decision support system which can help dermatologists in diagnosing of the skin lesions.

show abstract

The Balanced Accuracy and Its Posterior Distribution

Cited by 1,061 publications

References 5 publications

Assessment of spontaneous cardiovascular oscillations in Parkinson's disease

Assessment of spontaneous cardiovascular oscillations in Parkinson's disease

Bayesian environment representation, prediction, and criticality assessment for driver assistance systems

Skin Lesion Classification using Machine Learning Algorithms

Contact Info

Product

Resources

About