Automated analyses of electroencephalographic (EEG) signals acquired in naturalistic environments is becoming increasingly important in areas such as brain computer interfaces and behaviour science. However, the recorded EEG in such environments is often heavily contaminated by motion artifacts and eye movements. This poses new requirements on artifact handling. The objective of this paper is to present an automated EEG artifacts handling algorithm which will be used as a pre-processing step in a driver monitoring application. The algorithm, named ARTE (Automated aRTifacts handling in EEG), is based on wavelets, independent component analysis and hierarchical clustering. The algorithm is tested on a dataset obtained from a driver sleepiness study including 30 drivers and 540 30-minute 30-channel EEG recordings. The algorithm is evaluated by a clinical neurophysiologist, by quantitative criteria (signal quality index, mean square error, relative error and mean absolute error), and by demonstrating its usefulness as a preprocessing step in driver monitoring, here exemplified with driver sleepiness classification. All results are compared with a state of the art algorithm called FORCe. The quantitative and expert evaluation results show that the two algorithms are comparable and that both algorithms significantly reduce the impact of artifacts in recorded EEG signals. When artifact handling is used as a pre-processing step in driver sleepiness classification, the classification accuracy increased by 5% when using ARTE and by 2% when using FORCe. The advantage with ARTE is that it is data driven and does not rely on additional reference signals or manually defined thresholds, making it well suited for use in dynamic settings where unforeseen and rare artifacts are commonly encountered.

Assembly processes can be affected by various parameters, which is revealed by the measured geometrical characteristics (GCs) of the assembled parts deviating from the nominal values. Here, we propose a mixed-effect model (MEM) application for the purposes of analyzing variations in assembly cells, as well as for screening the input variables and characterization. MEMs make it possible to take into account statistical dependencies that originate from repeated measurements on the same assembly. The desirability functions approach was used to describe how to find corrective or control actions based on the fitted MEM. Objectives: To examine the usefulness of the MEM between the positions of the in-going parts as the input controllable variables and the measured GCs as the outputs. Methods: The data from 34 car frontal cross members (each measured three times) were experimentally collected in a laboratory environment by intentionally changing the positions of the in-going parts, assembling the parts, and subsequently measuring their GCs. A single MEM that completely describes the assembly process was fitted between the GCs and the positions of the in-going parts. Results: We present a modeling technique that can be used to establish which measured GCs are influenced by which controllable variables, and how this occurs. The fitted MEM shows evidence that the variability of some GCs changes over time. The natural variation in the system (i.e., unmodeled variations) is about two times larger than the variation between the assembled cross members. We also present two cases that demonstrate how to use the fitted MEM desirability functions to find corrective or control actions. Conclusion: MEMs are very useful tools for analyzing the assembly processes for car-body parts, which are nonlinear processes with multiple inputs and multiple correlated outputs. MEMs can potentially be applied in numerous industrial processes, since modern manufacturing plants measure all important process variables, which is the sole prerequisite for MEMs applications. 

The assembly quality is affected by various factors within which fixture variations are the most important. For that reason significant research on fixture variations has already been done. In this work we propose a linear mixed models (LMMs) application for the purpose of analyzing sources of variation in the fixture Objective: To estimate the strength of influences of different sources of variation on the control and assembly fixtures. The variables considered are: time, operator, default pin positions, shifts from the default pin positions . Methods: The data was collected through assembly and measurement for repeatability and experimental corrective actions. We use LMMs to model the relation between features measured on the assembled parts and the input variables of interest. The LMMs allow taking into account the correlation of observations contained in the dataset. We also use graphical data presentation methods to explore the data. Results: The expected results are the strengths of influences of the individual variables considered, and the pairwise interactions of between the variables, on the assembled parts variations.

This paper presents a novel approach to fault diagnosis applied to a stream of cases. The approach uses a combination of case-based reasoning and information fusion to do classification. The approach consists of two steps. First, we perform local anomaly detection on-board a machine to identify anomalous individual cases. Then, we monitor the stream of anomalous cases using a stream anomaly detector based on a sliding window approach. When the stream anomaly detector identifies an anomalous window, the anomalous cases in the window are classified using a CBR classifier. Thereafter, the individual classifications are aggregated into a composite case with a single prediction using a information fusion method. We compare three information fusion approaches: simple majority vote, weighted majority vote and Dempster-Shafer fusion. As baseline for comparison, we use the classification of the last identified anomalous case in the window as the aggregated prediction.

The rapid development of the medical field makes it impossible even for experts in the field to keep up with new treatments and experience. Already in 2010 all medical knowledge doubled in 3,5 years, to keep up to date with all development even in a narrow field is today far beyond human capacity. The need for decision support is increasingly important to ensure optimal treatment of patients, especially if patients are not "standard patients" matching a gold standard treatment. By ensuring confidentiality and collecting structured cases on a large scale will enable clinical decision support far beyond what is possible today and will be a major leap in healthcare. 

Today, usage of physiological sensor signals is essential in medical applications for diagnoses and classification of diseases. Clinicians often rely on information collected from several physiological sensor signals to diagnose a patient. However, sensor signals are mostly non-stationary and noisy, and single sensor signal could easily be contaminated by uncertain noises and interferences that could cause miscalculation of measurements and reduce clinical usefulness. Therefore, an apparent choice is to use multiple sensor signals that could provide more robust and reliable decision. Therefore, a physiological signal classification approach is presented based on sensor signal fusion and case-based reasoning. To classify Stressed and Relaxed individuals from physiological signals, data level and decision level fusion are performed and case-based reasoning is applied as classification algorithm. Five physiological sensor signals i.e., Heart Rate (HR), Finger Temperature (FT), Respiration Rate (RR), Carbon dioxide (CO2) and Oxygen Saturation (SpO2) are collected during the data collection phase. Here, data level fusion is performed using Multivariate Multiscale Entropy (MMSE) and extracted features are then used to build a case- library. Decision level fusion is performed on the features extracted using traditional time and frequency domain analysis. Case-Based Reasoning (CBR) is applied for the classification of the signals. The experimental result shows that the proposed system could classify Stressed or Relaxed individual 87.5% accurately compare to an expert in the domain. So, it shows promising result in the psychophysiological domain and could be possible to adapt this approach to other relevant healthcare systems.

This paper describes a generic framework for explaining the prediction of probabilistic machine learning algorithms using cases. The framework consists of two components: a similarity metric between cases that is defined relative to a probability model and an novel case-based approach to justifying the probabilistic prediction by estimating the prediction error using case-based reasoning. As basis for deriving similarity metrics, we define similarity in terms of the principle of interchangeability that two cases are considered similar or identical if two probability distributions, derived from excluding either one or the other case in the case base, are identical. Lastly, we show the applicability of the proposed approach by deriving a metric for linear regression, and apply the proposed approach for explaining predictions of the energy performance of households.

This paper describes a generic framework for explaining the prediction of a probabilistic classifier using preceding cases. Within the framework, we derive similarity metrics that relate the similarity between two cases to a probability model and propose a novel case-based approach to justifying a classification using the local accuracy of the most similar cases as a confidence measure. As a basis for deriving similarity metrics, we define similarity in terms of the principle of interchangeability that two cases are considered similar or identical if two probability distributions, derived from excluding either one or the other case in the case base, are identical. Thereafter, we evaluate the proposed approach for explaining the probabilistic classification of faults by logistic regression. We show that with the proposed approach, it is possible to find cases for which the used classifier accuracy is very low and uncertain, even though the predicted class has high probability.

This paper presents a generic approach to fault diagnosis of heavy duty machines that combines signal processing, statistics, machine learning, and case-based reasoning for on-board and off-board analysis. The used methods complement each other in that the on-board methods are fast and light-weight, while case-based reasoning is used off-board for fault diagnosis and for retrieving cases as support in manual decision mak- ing. Three major contributions are novel approaches to detecting clutch slippage, anomaly detection, and case-based diagnosis that is closely in- tegrated with the anomaly detection model. As example application, the proposed approach has been applied to diagnosing the root cause of clutch slippage in automatic transmissions.