Blind source separation and feature extraction in concurrent control charts pattern recognition: Novel analyses and a comparison of different methods

The most standard blind source separation (BSS) methods address the situation when a set of signals are available, e.g. from measurements, and all of them are linear memoryless combinations, with unknown coefficient values, of the same limited set of unknown source signals. BSS methods aim at estimating these unknown source signals and/or coefficients. This generic problem is e.g. faced in the field of Earth observation (where it is also called “unsupervised unmixing”), when considering the commonly used (over)simplified model of hyperspectral images. Each pixel of such an image has an associated reflectance spectrum derived from measurements, which is defined by the fraction of sunlight power reflected by the corresponding Earth surface at each wavelength. Each source signal is then the single reflectance spectrum associated with one of the classes of pure materials which are present in the region of Earth corresponding to the overall considered hyperspectral image. Besides, the associated coefficients define the surfaces on Earth covered with each of these pure materials in each sub-region corresponding to one pixel of the considered image. However, real hyperspectral data e.g. obtained in urban areas have a much more complex structure than the above basic model: each class of pure materials (e.g. roof tiles, trees or asphalt) has so-called spectral or intra-class variability, i.e. it yields a somewhat different spectral component in each pixel of the image. In this complex framework, this chapter shows that Principal Component Analysis (PCA) and its proposed extension are of high interest at three stages of our investigation. First, PCA allows us to analyze the above-mentioned spectral variability of real high-dimensional hyperspectral data and to propose an extended data model which is suited to these complex data. We then develop a versatile extension of BSS methods based on Nonnegative Matrix Factorization, which adds the capability to handle arbitrary forms of intra-class variability by transposing PCA concepts to this original version of the BSS tramework. Finally, PCA again proves to be very well suited to analyzing the high-dimensional data obtained as the result of the proposed BSS method.

Upgrading Sparse NMF algorithm for blind source separation through Adaptive Parameterized Hybrid Kernel based approach

This paper presents a convex geometry (CG)-based method for blind separation of nonnegative sources. First, the unaccessible source matrix is normalized to be column-sum-to-one by mapping the available observation matrix. Then, its zero-samples are found by searching the facets of the convex hull spanned by the mapped observations. Considering these zero-samples, a quadratic cost function with respect to each row of the unmixing matrix, together with a linear constraint in relation to the involved variables, is proposed. Upon which, an algorithm is presented to estimate the unmixing matrix by solving a classical convex optimization problem. Unlike the traditional blind source separation (BSS) methods, the CG-based method does not require the independence assumption, nor the uncorrelation assumption. Compared with the BSS methods that are specifically designed to distinguish between nonnegative sources, the proposed method requires a weaker sparsity condition. Provided simulation results illustrate the performance of our method.

The monitoring of processes is a vital mechanism for ensuring that such processes remain safe and under control. The present research aims to solve problems associated with correlated data by applying the Box-Jenkins method integrated with statistical process control (SPC) tools, namely the Shewhart chart, the moving average chart, the cumulative sum (CUSUM) chart, and the exponentially weighted moving-average (EWMA) chart. The efficiency of the four SPC tools was also compared in terms of the false alarm rate (FAR) and the missed detection rate (MDR). The findings indicated that the EWMA chart was the most effective in detecting anomaly, the Shewhart chart and the moving average chart produced high MDR, and the CUSUM chart suffered the highest FAR.

A statistical weld process monitoring system is described. Using data collected while welding, the welding statistical process control (SPC) tool provides weld process quality control by implementing techniques of data trending analysis, tolerance analysis, and sequential analysis. The SPC system computes the mean, standard deviation, and range of each of the parameters sampled by the data collection system. Changes in the mean, standard deviation, and range are displayed using control (or trend) charts. The control chart displays a function of a parameter with respect to the ordering of the weld records (for a single weld) or weld number (for multiple welds). The SPC tool also permits plotting tolerance charts of the mean, standard deviation, and range for each of the sampled parameters. The tolerance chart is plotted versus the record number (or weld number) and consists of a vertical line for each record (or weld number) showing the minimum and maximum value of that parameter for that record (or weld number). The upper control limit (UCL), lower control limit (LCL), and nominal value may also be displayed on the tolerance chart printout. The SPC also performs sequential analysis, which allows the user to examine the process as it goes along, which in turn may permit the user to locate a possible change in the process before it goes out of control. Work directed toward developing an expert interpreter of the voluminous statistical output generated by the SPC is also described.

In the recent past, blind source separation (BSS) algorithms using multivariate statistical data analysis technique have been successfully used for source identification and separation in the field of biomedical and statistical signal processing. Recently numbers of different BSS techniques have been developed. With BSS methods being the feasible method for source separation and decomposition of biosignals, it is important to compare the different techniques and determine the most suitable method for the applications. This paper presents the performance of five BSS algorithms (SOBI, TDSEP, FastICA, JADE and Infomax) for decomposition of sEMG to identify subtle finger movements. It is observed that BSS algorithms based on second-order statistics (SOBI and TDSEP) gives better performance compared to algorithms based on higher-order statistics (FastICA, JADE and infomax).

Blind source separation has found applications in various areas including biomedical signal processing and genomic signal processing. Often, blind source separation is solved via independent component analysis (ICA) by assuming and utilizing mutual independence among source signals. However, in bio-signal and genomic signal processing, the assumption of independence is often untrue, and the performance of the ICA approach is not as good. Much effort has been devoted to searching alternative approaches to blind source separation without the independence assumption. One idea known as network component analysis (NCA) is developed to identify the underlying regulatory signals of transcription factors in the gene regulatory network. In this paper we show that NCA is a general method for blind source separation using a priori information on the mixing matrix. An alternative proof of identifiability using NCA is proposed and a novel method to solve the problem is developed. Validation is made through computer simulations.

Moving towards process-based radiotherapy quality assurance using statistical process control

Nowadays, statistical process control (SPC) is extensively utilized in manufacturing companies. Control charts are the foremost tool in SPC because they can timely detect assignable causes that influence the quality of the production process. Mostly, the control charts are used to monitor the process variability and the process mean. However, when the mean changes but is appraised as in-control and the standard deviation is linear to the mean, the Coefficient of Variation (CV) chart is suitable for evaluating process variability. In this study, we suggest a Generally Weighted Moving Coefficient of Variation (GWMCV) control chart using three-parameter logarithmic transformations which parameter weight for each time (ω) is more flexible than the Exponentially Weighted Moving Coefficient of Variation (EWMCV) control chart that has ω = 1. The Simulation studies show that when parameter weight 0.6≤ω≤0.9,GWMCV chart is more sensitive than EWMCV chart. As a result, the flexible parameter weights increase the diagnosis ability of the control chart. The chart is also applied to monitor the process production of NPK fertilizer. The result shows that proposed chart detect out-of-control signal rapidly than the EWMCV.

Control charts (CCs) are one of the main tools in Statistical Process Control that have been widely adopted in manufacturing sectors as an effective strategy for malfunction detection throughout the previous decades. Measurement errors (M.E's) are involved in the quality characteristic of interest, which can effect the CC's performance. The authors explored the impact of a linear model with additive covariate M.E on the multivariate cumulative sum (CUSUM) CC for a specific kind of data known as compositional data (CoDa). The average run length is used to assess the performance of the proposed chart. The results indicate that M.E's significantly affects the multivariate CUSUM-CoDa CCs. The authors have used the Markov chain method to study the impact of different involved parameters using six different cases for the variance-covariance matrix (VCM) (i.e., uncorrelated with equal variances, uncorrelated with unequal variances, positively correlated with equal variances, positively correlated with unequal variances, negatively correlated with equal variances and negatively correlated with unequal variances). The authors concluded that the error VCM has a negative impact on the performance of the multivariate CUSUM-CoDa CC, as the increases with an increase in the value of the error VCM. The subgroup size m and powering operator b positively impact the proposed CC, as the decreases with an increase in m or b. The number of variables p also has a negative impact on the performance of the proposed CC, as the values of increase with an increase in p. For the implementation of the proposal, two illustrated examples have been reported for multivariate CUSUM-CoDa CCs in the presence of M.E's. One deals with the manufacturing process of uncoated aspirin tablets, and the other is based on monitoring the machines involved in the muesli manufacturing process.

A control chart is a tool for statistical process control (SPC) used to determine variations in manufacturing processes resulting from common or assignable causes. The presence of unnatural patterns in control charts is an indication that the process has been influenced by assignable causes, and corrective actions must be taken. However, the assumption of uncorrelated or independent observations is not suitable for all of the characteristics pertaining to a product. This study presents a method with which to improve the accuracy of pattern recognition in control charts, through the autocorrelation of product characteristics. Particularly, we developed an artificial algorithm-based machine learning model to recognize unnatural patterns and processes in AR(1) simultaneously. The proposed method integrates an artificial immune system and support vector machine within a recognition system. This study evaluated the accuracy with which four patterns could be recognized. The four patterns include trends, sudden shifts, cyclic patterns, and normal patterns. Identifying unnatural patterns can greatly narrow the range of possible causes that must be investigated, thereby speeding up the diagnostic processes. Key words: Pattern recognition, artificial immune algorithm, support vector machine, autocorrelation.

Drive system is one of most important key equipment to guarantee safe and stable operation in locomotive. With time variation, unpredictability and nonstationary, fault source of drive system is not obtained by traditional fault diagnosis method. Blind source separation is a kind of method on source signals separation under transmission channel unknown instance. The method of Blind source separation based on variable metric empirical mode decomposition is proposed. Intrinsic mode function is built, redundancy factors are reduced, and recurrent neural network is used to adaptive blind separation. The method is verified by data analysis of on-line measuring. The results show that separation efficiency is improved and unaffected with iteration time in the process of fault information separation, which will be better for further fundamental research and provide technique support for the locomotive.

Event-related potentials (ERPs) are a superposition of electric potential differences generated by neurophysiological activity associated with psychophysical events. Spatiotemporal dissociation of these signal sources can supplement conventional ERP analysis and improve source localization. However, results from established source separation methods applied to ERPs can be challenging to interpret. Hence, we have developed a recurrent neural network (RNN) method for blind source separation. The RNN transforms input step pulse signals representing events into corresponding ERP difference waveforms. Source waveforms are obtained from penultimate layer units and scalp maps are obtained from feed-forward output layer weights that project these source waveforms onto EEG electrode amplitudes. An interpretable, sparse source representation is achieved by incorporating L1 regularization of signals obtained from the penultimate layer of the network during training. This RNN method was applied to four ERP difference waveforms (MMN, N170, N400, P3) from the open-access ERP CORE database, and independent component analysis (ICA) was applied to the same data for comparison. The RNN decomposed these ERPs into eleven spatially and temporally separate sources that were less noisy, tended to be more ERP-specific, and were less similar to each other than ICA-derived sources. The RNN sources also had less ambiguity between source waveform amplitude, scalp potential polarity, and equivalent current dipole orientation than ICA sources. In conclusion, the proposed RNN blind source separation method can be effectively applied to grand-average ERP difference waves and holds promise for further development as a computational model of event-related neural signals.

We present a method of blind source separation (BSS) for speech signals using a complex extension of infinite sparse factor analysis (ISFA) in the frequency domain. Our method is robust against delayed signals that usually occur in real environments, such as reflections, short-time reverberations, and time lags of signals arriving at microphones. ISFA is a conventional non-parametric Bayesian method of BSS, which has only been applied to time domain signals because it can only deal with real signals. Our method uses complex normal distributions to estimate source signals and mixing matrix. Experimental results indicate that our method outperforms the conventional ISFA in the average signal-to-distortion ratio (SDR).

Conventionally, there is a method to use auscultation when medical doctors diagnose a disease from heartbeats with heart noise. However, the respiratory sound and the heart sound are mixed. Further, it is difficult for the desired heart sound to be heard. For solving this problem, we propose a method to separate respiratory sound and heart sound based on blind source separation (BSS). In BSS method, we use a non-linear function corresponding to the probability distribution of source signals in order to separate. A distribution of the respiratory sound is a high-order sub-Gaussian distribution close to uniform distribution, a distribution of the heart sound is a super Gaussian distribution. In the conventional method, there is no non-linear function corresponding to both distribution. Therefore, in this paper, in order to separate the respiratory sound and the heart sound, we propose new non-linear function accurately separated from super Gaussian distribution through sub-Gaussian distribution to uniform distribution. Then, we build the BSS method based on the proposed non-linear function. We evaluate the proposed method by computer simulation.