Search

In machine learning, data augmentation is the process of generating synthetic data samples that will be used to train the model to improve the performance of the machine learning model. Data augmentation has been shown to improve the generalization capabilities of models and is particularly popular in the field of computer vision, which is to deal with image data. In contrast, data augmentation is less widely used in the field of time-series data, such as time-series classification, than in computer vision. This is because time-series data is particularly vulnerable to the transformation of data that occurs while performing data augmentation. For example, flipping or rotating images to augment image data does not significantly undermine the meaning of the original data. However, in the case of time-series data, it is likely to distort the meaning of the data, and it is difficult to identify whether or not it has altered. Previously proposed time-series data augmentation methods performed well in many fields, but often did not consider trend information of time-series data such as slicing or reordering the original time-series. In this paper, we propose a time-series data augmentation method based on interpolation. The proposed method is robust against the impairment of trend information of the original time-series and has the advantage of not high complexity. To evaluate the performance of the proposed method, we experimented with time-series datasets from the UCR archive and showed that the performance of the model could be improved.

Process monitoring at industrial sites contributes to system stability by detecting and diagnosing unexpected changes in a system. Today, as the infrastructure of industrial sites is advanced due to the development of communication technology, various and vast amounts of data are generated, and the importance of a methodology to effectively monitor these data to diagnose a system is increasing daily. As a deep neural network-based methodology can effectively extract information from a large amount of data, methods have been proposed to monitor processes using this methodology to detect any system abnormalities. Neural network-based process monitoring is effective in detecting anomalies but has difficulty in diagnosing due to the limitations of the black-box model. Therefore, this paper proposes a process monitoring framework that can detect and diagnose anomalies. The proposed framework performs post-processing based on the class activation map to perform the diagnosis of data that are considered outliers. To verify the performance of the proposed method, experiments were conducted using industrial public motor datasets, demonstrating that the proposed method can effectively detect and diagnose abnormalities.

We report a novel peak sorting method for the two-dimensional gas chromatography/time-of-flight mass spectrometry (GC x GC/TOF-MS) system. The objective of peak sorting is to recognize peaks from the same metabolite occurring in different samples from thousands of peaks detected in the analytical procedure. The developed algorithm is based on the fact that the chromatographic peaks for a given analyte have similar retention times in all of the chromatograms. Raw instrument data are first processed by ChromaTOF (Leco) software to provide the peak tables. Our algorithm achieves peak sorting by utilizing the first- and second-dimension retention times in the peak tables and the mass spectra generated during the process of electron impact ionization. The algorithm searches the peak tables for the peaks generated by the same type of metabolite using several search criteria. Our software also includes options to eliminate non-target peaks from the sorting results, e.g., peaks of contaminants. The developed software package has been tested using a mixture of standard metabolites and another mixture of standard metabolites spiked into human serum. Manual validation demonstrates high accuracy of peak sorting with this algorithm.

A class of interconnected neural networks composed of generalized Brain-State-in-a-Box (gBSB) neural subnetworks is considered. Interconnected gBSB neural network architectures are proposed along with their stability conditions. The design of the interconnected neural networks is reduced to the problem of solving linear matrix inequalities (LMIs) to determine the interconnection parameters. A method for solving LMIs is devised generating the solutions that, in general, are further away from zero than the corresponding solutions obtained using MATLAB's LMI toolbox, thus resulting in stronger interconnections between the subnetworks. The proposed architectures are then used to construct neural associative memories. Simulations are performed to illustrate the results obtained.

Systems biology aims to understand biological systems on a comprehensive scale, such that the components that make up the whole are connected to one another and work in harmony. As a major component of systems biology, differential proteomics studies the differences between distinct but related proteomes such as normal versus diseased cells and diseased versus treated cells. High throughput mass spectrometry (MS) based analytical platforms are widely used in differential proteomics (Domon, 2006; Fenselau, 2007). As a common practice, the proteome is usually digested into peptides first. The peptide mixture is then separated using multidimensional liquid chromatography (MDLC) and is finally subjected to MS for further analysis. Thousands of mass spectra are generated in a single experiment. Discovering the significantly changed proteins from millions of peaks involves mass informatics. This paper introduces data mining steps used in mass informatics, and concludes with a descriptive examination of concepts, trends and challenges in this rapidly expanding field.

Overview We have developed the Proteome Discovery Pipeline, a stand-alone bioinformatics platform used for LC/MS data analysis and biomarker discovery. Data is processed in a series of self-contained analytical steps using modules that are controlled by a graphical user interface. The user interface was developed in Visual C++ 6.0 and provides a multi-threaded, tabbed user interface with each tab representing a step in the analysis process. Modules included are spectrum deconvolution, alignment, normalization, significance tests and pattern recognition. Modules consist of applications developed in C++ and the R scripting language, which are called as external processes from the GUI using inputted parameters. Molecular correlation analysis can be viewed interactively using SysNet. Figure 1 shows the architecture of the Proteome Discovery Pipeline.

The majority of metabolomic studies used in ecotoxicology have implemented (1)H NMR analysis. Despite constant improvement, major limitations of NMR-based techniques include relatively low sensitivity that results in an examination of a limited number of metabolites. An alternative approach is the use of liquid or gas chromatography (GC) for separation of metabolites and mass spectrometry (MS) for their quantification and identification. The objective of our study was to develop a two dimensional GC coupled with time of flight MS (GCxGC/TOF-MS) coupled with multivariate analysis to compare metabolite profiles of Diporeia under different environmental conditions. We compared metabolite profiles between Diporeia collected from Lake Michigan (declining populations) to those residing in Lake Superior (stable populations), and also between Diporeia exposed to a chemical stressor (atrazine) and controls. Overall, 76 and 302 total metabolites were detected from the lake comparison and atrazine studies, respectively. Many of the identified metabolites included fatty acids, amino acids, and hydrocarbons. Furthermore, we observed unique and almost non-overlapping metabolite profiles in both studies. In conclusion, we established the feasibility of using GCxGC/TOF-MS for detecting metabolites as well as developed software to align and merge chromatographic peaks to compare metabolite differences between invertebrate groups sampled under different environmental conditions. This ability to detect unique metabolite profiles under different environmental conditions will increase our undertsanding on the physiological processes and whole-organism reponses occuring as a result of exposure to different environmental stressors.

Motivation: The still emerging combination of technologies that enable description and characterization of all expressed proteins in a biological system is known as proteomics. Although many separation and analysis technologies have been employed in proteomics, it remains a challenge to predict peptide behavior during separation processes. New informatics tools are needed to model the experimental analysis method that will allow scientists to predict peptide separation and assist with required data mining steps, such as protein identification. Results: We developed a software package to predict the separation of peptides in strong anion exchange (SAX) chromatography using artificial neural network based pattern classification techniques. A multi-layer perceptron is used as a pattern classifier and it is designed with feature vectors extracted from the peptides so that the classification error is minimized. A genetic algorithm is employed to train the neural network. The developed system was tested using 14 protein digests, and the sensitivity analysis was carried out to investigate the significance of each feature. Availability: The software and testing results can be downloaded from . Contact: zhang100@purdue.edu