Search results for: Multivariate Data Analysis

Authors: Francisco Aparisi, José Sanz

Abstract:

Multivariate quality control charts show some advantages to monitor several variables in comparison with the simultaneous use of univariate charts, nevertheless, there are some disadvantages. The main problem is how to interpret the out-ofcontrol signal of a multivariate chart. For example, in the case of control charts designed to monitor the mean vector, the chart signals showing that it must be accepted that there is a shift in the vector, but no indication is given about the variables that have produced this shift. The MEWMA quality control chart is a very powerful scheme to detect small shifts in the mean vector. There are no previous specific works about the interpretation of the out-of-control signal of this chart. In this paper neural networks are designed to interpret the out-of-control signal of the MEWMA chart, and the percentage of correct classifications is studied for different cases.

Keywords: Multivariate quality control, Artificial Intelligence, Neural Networks, Computer Applications

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2465

Authors: L. Melecký

Abstract:

Measurement of competitiveness between countries or regions is an important topic of many economic analysis and scientific papers. In European Union (EU), there is no mainstream approach of competitiveness evaluation and measuring. There are many opinions and methods of measurement and evaluation of competitiveness between states or regions at national and European level. The methods differ in structure of using the indicators of competitiveness and ways of their processing. The aim of the paper is to analyze main sources of competitive potential of the EU Member States with the help of Factor analysis (FA) and to classify the EU Member States to homogeneous units (clusters) according to the similarity of selected indicators of competitiveness factors by Cluster analysis (CA) in reference years 2000 and 2011. The theoretical part of the paper is devoted to the fundamental bases of competitiveness and the methodology of FA and CA methods. The empirical part of the paper deals with the evaluation of competitiveness factors in the EU Member States and cluster comparison of evaluated countries by cluster analysis. 

Keywords: Competitiveness, cluster analysis, EU, factor analysis, multivariate methods.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2001

Authors: Alexander Efremov

Abstract:

A system for market identification (SMI) is presented. The resulting representations are multivariable dynamic demand models. The market specifics are analyzed. Appropriate models and identification techniques are chosen. Multivariate static and dynamic models are used to represent the market behavior. The steps of the first stage of SMI, named data preprocessing, are mentioned. Next, the second stage, which is the model estimation, is considered in more details. Stepwise linear regression (SWR) is used to determine the significant cross-effects and the orders of the model polynomials. The estimates of the model parameters are obtained by a numerically stable estimator. Real market data is used to analyze SMI performance. The main conclusion is related to the applicability of multivariate dynamic models for representation of market systems.

Keywords: market identification, dynamic models, stepwise regression.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1576

Authors: Sachin Kumar, Vasilis Sotiris, Michael Pecht

Abstract:

With increasing complexity in electronic systems there is a need for system level anomaly detection and fault isolation. Anomaly detection based on vector similarity to a training set is used in this paper through two approaches, one the preserves the original information, Mahalanobis Distance (MD), and the other that compresses the data into its principal components, Projection Pursuit Analysis. These methods have been used to detect deviations in system performance from normal operation and for critical parameter isolation in multivariate environments. The study evaluates the detection capability of each approach on a set of test data with known faults against a baseline set of data representative of such “healthy" systems.

Keywords: Mahalanobis distance, Principle components, Projection pursuit, Health assessment, Anomaly.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1640

Authors: B. Nassar, W. Hussein, M. Mokhtar

Abstract:

The critical concern of satellite operations is to ensure the health and safety of satellites. The worst case in this perspective is probably the loss of a mission, but the more common interruption of satellite functionality can result in compromised mission objectives. All the data acquiring from the spacecraft are known as Telemetry (TM), which contains the wealth information related to the health of all its subsystems. Each single item of information is contained in a telemetry parameter, which represents a time-variant property (i.e. a status or a measurement) to be checked. As a consequence, there is a continuous improvement of TM monitoring systems to reduce the time required to respond to changes in a satellite's state of health. A fast conception of the current state of the satellite is thus very important to respond to occurring failures. Statistical multivariate latent techniques are one of the vital learning tools that are used to tackle the problem above coherently. Information extraction from such rich data sources using advanced statistical methodologies is a challenging task due to the massive volume of data. To solve this problem, in this paper, we present a proposed unsupervised learning algorithm based on Principle Component Analysis (PCA) technique. The algorithm is particularly applied on an actual remote sensing spacecraft. Data from the Attitude Determination and Control System (ADCS) was acquired under two operation conditions: normal and faulty states. The models were built and tested under these conditions, and the results show that the algorithm could successfully differentiate between these operations conditions. Furthermore, the algorithm provides competent information in prediction as well as adding more insight and physical interpretation to the ADCS operation.

Keywords: Space telemetry monitoring, multivariate analysis, PCA algorithm, space operations.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2026

Authors: O. Ikpotokin

Abstract:

In any production process, every product is aimed to attain a certain standard, but the presence of assignable cause of variability affects our process, thereby leading to low quality of product. The ability to identify and remove this type of variability reduces its overall effect, thereby improving the quality of the product. In case of a univariate control chart signal, it is easy to detect the problem and give a solution since it is related to a single quality characteristic. However, the problems involved in the use of multivariate control chart are the violation of multivariate normal assumption and the difficulty in identifying the quality characteristic(s) that resulted in the out of control signals. The purpose of this paper is to examine the use of non-parametric control chart (the bootstrap approach) for obtaining control limit to overcome the problem of multivariate distributional assumption and the p-value method for detecting out of control signals. Results from a performance study show that the proposed bootstrap method enables the setting of control limit that can enhance the detection of out of control signals when compared, while the p-value method also enhanced in identifying out of control variables.

Keywords: Bootstrap control limit, p-value method, out-of-control signals, p-value, quality characteristics.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 971

Authors: S. Barbosa, M. Pinto, J. A. Almeida, E. Carvalho, C. Diamantino

Abstract:

The aim of this work was to test a methodology able to generate spatial-temporal maps that can synthesize simultaneously the trends of distinct hydrochemical indicators in an old radium-uranium tailings dam deposit. Multidimensionality reduction derived from principal component analysis and subsequent data aggregation derived from clustering analysis allow to identify distinct hydrochemical behavioral profiles and generate synthetic evolutionary hydrochemical maps.

Keywords: Contamination plume migration, K-means of PCA scores, groundwater and mine water monitoring, spatial-temporal hydrochemical trends.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 568

Authors: P. V. Pramila, V. Mahesh

Abstract:

Pulmonary Function Tests are important non-invasive diagnostic tests to assess respiratory impairments and provides quantifiable measures of lung function. Spirometry is the most frequently used measure of lung function and plays an essential role in the diagnosis and management of pulmonary diseases. However, the test requires considerable patient effort and cooperation, markedly related to the age of patients resulting in incomplete data sets. This paper presents, a nonlinear model built using Multivariate adaptive regression splines and Random forest regression model to predict the missing spirometric features. Random forest based feature selection is used to enhance both the generalization capability and the model interpretability. In the present study, flow-volume data are recorded for N= 198 subjects. The ranked order of feature importance index calculated by the random forests model shows that the spirometric features FVC, FEF25, PEF, FEF25-75, FEF50 and the demographic parameter height are the important descriptors. A comparison of performance assessment of both models prove that, the prediction ability of MARS with the `top two ranked features namely the FVC and FEF25 is higher, yielding a model fit of R2= 0.96 and R2= 0.99 for normal and abnormal subjects. The Root Mean Square Error analysis of the RF model and the MARS model also shows that the latter is capable of predicting the missing values of FEV1 with a notably lower error value of 0.0191 (normal subjects) and 0.0106 (abnormal subjects) with the aforementioned input features. It is concluded that combining feature selection with a prediction model provides a minimum subset of predominant features to train the model, as well as yielding better prediction performance. This analysis can assist clinicians with a intelligence support system in the medical diagnosis and improvement of clinical care.

Keywords: FEV1, Multivariate Adaptive Regression Splines Pulmonary Function Test, Random Forest.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 3692

Authors: Nazif Çalış, Murat Erişoğlu, Hamza Erol, Tayfun Servi

Abstract:

In the recent works related with mixture discriminant analysis (MDA), expectation and maximization (EM) algorithm is used to estimate parameters of Gaussian mixtures. But, initial values of EM algorithm affect the final parameters- estimates. Also, when EM algorithm is applied two times, for the same data set, it can be give different results for the estimate of parameters and this affect the classification accuracy of MDA. Forthcoming this problem, we use Self Organizing Mixture Network (SOMN) algorithm to estimate parameters of Gaussians mixtures in MDA that SOMN is more robust when random the initial values of the parameters are used [5]. We show effectiveness of this method on popular simulated waveform datasets and real glass data set.

Keywords: Self Organizing Mixture Network, MixtureDiscriminant Analysis, Waveform Datasets, Glass Identification, Mixture of Multivariate Normal Distributions

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1484

Authors: Tahseen A. Jilani, S. M. Aqil Burney, C. Ardil

Abstract:

In this paper, we have presented a new multivariate fuzzy time series forecasting method. This method assumes mfactors with one main factor of interest. History of past three years is used for making new forecasts. This new method is applied in forecasting total number of car accidents in Belgium using four secondary factors. We also make comparison of our proposed method with existing methods of fuzzy time series forecasting. Experimentally, it is shown that our proposed method perform better than existing fuzzy time series forecasting methods. Practically, actuaries are interested in analysis of the patterns of causalities in road accidents. Thus using fuzzy time series, actuaries can define fuzzy premium and fuzzy underwriting of car insurance and life insurance for car insurance. National Institute of Statistics, Belgium provides region of risk classification for each road. Thus using this risk classification, we can predict premium rate and underwriting of insurance policy holders.

Keywords: Average forecasting error rate (AFER), Fuzziness offuzzy sets Fuzzy, If-Then rules, Multivariate fuzzy time series.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2450

Authors: S. Aouabdi, M. Taibi

Abstract:

This paper presents powerful techniques for the development of a new monitoring method based on multi-scale entropy (MSE) in order to characterize the behaviour of the concentrations of different gases present in the synthesis of Ammonia and soft-sensor based on Principal Component Analysis (PCA).

Keywords: Ammonia synthesis, concentrations of different gases, soft sensor, multi-scale entropy, multivariate statistics.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2104

Authors: Entisar A. Elgmati

Abstract:

Recurrent event data is a special type of multivariate survival data. Dynamic and frailty models are one of the approaches that dealt with this kind of data. A comparison between these two models is studied using the empirical standard deviation of the standardized martingale residual processes as a way of assessing the fit of the two models based on the Aalen additive regression model. Here we found both approaches took heterogeneity into account and produce residual standard deviations close to each other both in the simulation study and in the real data set.

Keywords: Dynamic, frailty, misspecification, recurrent events.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2319

Authors: Pascal Mwenge, Tumisang Seodigeng

Abstract:

The combination of world population and the third industrial revolution led to high demand for fuels. On the other hand, the decrease of global fossil 8fuels deposits and the environmental air pollution caused by these fuels has compounded the challenges the world faces due to its need for energy. Therefore, new forms of environmentally friendly and renewable fuels such as biodiesel are needed. The primary analytical techniques for methanolysis yield monitoring have been chromatography and spectroscopy, these methods have been proven reliable but are more demanding, costly and do not provide real-time monitoring. In this work, the in situ monitoring of biodiesel from sunflower oil using FTIR (Fourier Transform Infrared) has been studied; the study was performed using EasyMax Mettler Toledo reactor equipped with a DiComp (Diamond) probe. The quantitative monitoring of methanolysis was performed by building a quantitative model with multivariate calibration using iC Quant module from iC IR 7.0 software. 15 samples of known concentrations were used for the modelling which were taken in duplicate for model calibration and cross-validation, data were pre-processed using mean centering and variance scale, spectrum math square root and solvent subtraction. These pre-processing methods improved the performance indexes from 7.98 to 0.0096, 11.2 to 3.41, 6.32 to 2.72, 0.9416 to 0.9999, RMSEC, RMSECV, RMSEP and R2Cum, respectively. The R² value of 1 (training), 0.9918 (test), 0.9946 (cross-validation) indicated the fitness of the model built. The model was tested against univariate model; small discrepancies were observed at low concentration due to unmodelled intermediates but were quite close at concentrations above 18%. The software eliminated the complexity of the Partial Least Square (PLS) chemometrics. It was concluded that the model obtained could be used to monitor methanol of sunflower oil at industrial and lab scale.

Keywords: Biodiesel, calibration, chemometrics, FTIR, methanolysis, multivariate analysis, transesterification.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 870

Authors: Jonathan D. Morales M., Ramón Silva R.

Abstract:

The purpose of this project is to carry out an analysis and determine the profile of actual lean manufacturing processes in the Metropolitan Area of Bucaramanga. Through the analysis of qualitative and quantitative variables it was possible to establish how these manufacturers develop production practices that ensure their competitiveness and productivity in the market. In this study, a random sample of metallurgic and wrought iron companies was applied, following which a quantitative focus and analysis was used to formulate a qualitative methodology for measuring the level of lean manufacturing procedures in the industry. A qualitative evaluation was also carried out through a multivariate analysis using the Numerical Taxonomy System (NTSYS) program which should allow for the determination of Lean Manufacturing profiles. Through the results it was possible to observe how the companies in the sector are doing with respect to Lean Manufacturing Practices, as well as identify the level of management that these companies practice with respect to this topic. In addition, it was possible to ascertain that there is no one dominant profile in the sector when it comes to Lean Manufacturing. It was established that the companies in the metallurgic and wrought iron industry show low levels of Lean Manufacturing implementation. Each one carries out diverse actions that are insufficient to consolidate a sectoral strategy for developing a competitive advantage which enables them to tie together a production strategy.

Keywords: Lean manufacturing, metallurgic industry, production line management, productivity.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1835

Authors: Reza Nadimi, Fariborz Jolai

Abstract:

This article combines two techniques: data envelopment analysis (DEA) and Factor analysis (FA) to data reduction in decision making units (DMU). Data envelopment analysis (DEA), a popular linear programming technique is useful to rate comparatively operational efficiency of decision making units (DMU) based on their deterministic (not necessarily stochastic) input–output data and factor analysis techniques, have been proposed as data reduction and classification technique, which can be applied in data envelopment analysis (DEA) technique for reduction input – output data. Numerical results reveal that the new approach shows a good consistency in ranking with DEA.

Keywords: Effectiveness, Decision Making, Data EnvelopmentAnalysis, Factor Analysis

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2384

Authors: S. Soommat, S. Patamatamkul, T. Prempridi, M. Sritulyachot, P. Ineure, S. Yimman

Abstract:

Currently, slider process of Hard Disk Drive Industry become more complex, defective diagnosis for yield improvement becomes more complicated and time-consumed. Manufacturing data analysis with data mining approach is widely used for solving that problem. The existing mining approach from combining of the KMean clustering, the machine oriented Kruskal-Wallis test and the multivariate chart were applied for defective diagnosis but it is still be a semiautomatic diagnosis system. This article aims to modify an algorithm to support an automatic decision for the existing approach. Based on the research framework, the new approach can do an automatic diagnosis and help engineer to find out the defective factors faster than the existing approach about 50%.

Keywords: Slider process, Defective diagnosis and Data mining.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1167

Authors: M. Heikkinen, A. Sarpola, H. Hellman, J. Rämö, Y. Hiltunen

Abstract:

Independent component analysis (ICA) is a computational method for finding underlying signals or components from multivariate statistical data. The ICA method has been successfully applied in many fields, e.g. in vision research, brain imaging, geological signals and telecommunications. In this paper, we apply the ICA method to an analysis of mass spectra of oligomeric species emerged from aluminium sulphate. Mass spectra are typically complex, because they are linear combinations of spectra from different types of oligomeric species. The results show that ICA can decomposite the spectral components for useful information. This information is essential in developing coagulation phases of water treatment processes.

Keywords: Independent component analysis, massspectroscopy, water treatment, aluminium sulphate.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2335

Authors: Fabrice Poirion, Benedicte Puig

Abstract:

Mathematical justifications are given for a simulation technique of multivariate nonGaussian random processes and fields based on Rosenblatt-s transformation of Gaussian processes. Different types of convergences are given for the approaching sequence. Moreover an original numerical method is proposed in order to solve the functional equation yielding the underlying Gaussian process autocorrelation function.

Keywords: Simulation, nonGaussian, random field, multivariate, stochastic process.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1787

Authors: Andreas Theissler, Ian Dear

Abstract:

In the automotive industry test drives are being conducted during the development of new vehicle models or as a part of quality assurance of series-production vehicles. The communication on the in-vehicle network, data from external sensors, or internal data from the electronic control units is recorded by automotive data loggers during the test drives. The recordings are used for fault analysis. Since the resulting data volume is tremendous, manually analysing each recording in great detail is not feasible. This paper proposes to use machine learning to support domainexperts by preventing them from contemplating irrelevant data and rather pointing them to the relevant parts in the recordings. The underlying idea is to learn the normal behaviour from available recordings, i.e. a training set, and then to autonomously detect unexpected deviations and report them as anomalies. The one-class support vector machine “support vector data description” is utilised to calculate distances of feature vectors. SVDDSUBSEQ is proposed as a novel approach, allowing to classify subsequences in multivariate time series data. The approach allows to detect unexpected faults without modelling effort as is shown with experimental results on recordings from test drives.

Keywords: Anomaly detection, fault detection, test drive analysis, machine learning.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2435

Authors: Hyun-Woo Cho

Abstract:

Detection of incipient abnormal events is important to improve safety and reliability of machine operations and reduce losses caused by failures. Improper set-ups or aligning of parts often leads to severe problems in many machines. The construction of prediction models for predicting faulty conditions is quite essential in making decisions on when to perform machine maintenance. This paper presents a multivariate calibration monitoring approach based on the statistical analysis of machine measurement data. The calibration model is used to predict two faulty conditions from historical reference data. This approach utilizes genetic algorithms (GA) based variable selection, and we evaluate the predictive performance of several prediction methods using real data. The results shows that the calibration model based on supervised probabilistic principal component analysis (SPPCA) yielded best performance in this work. By adopting a proper variable selection scheme in calibration models, the prediction performance can be improved by excluding non-informative variables from their model building steps.

Keywords: Prediction, operation monitoring, on-line data, nonlinear statistical methods, empirical model.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1627

Authors: Hyun-Woo Cho

Abstract:

Real-time or in-line process monitoring frameworks are designed to give early warnings for a fault along with meaningful identification of its assignable causes. In artificial intelligence and machine learning fields of pattern recognition various promising approaches have been proposed such as kernel-based nonlinear machine learning techniques. This work presents a kernel-based empirical monitoring scheme for batch type production processes with small sample size problem of partially unbalanced data. Measurement data of normal operations are easy to collect whilst special events or faults data are difficult to collect. In such situations, noise filtering techniques can be helpful in enhancing process monitoring performance. Furthermore, preprocessing of raw process data is used to get rid of unwanted variation of data. The performance of the monitoring scheme was demonstrated using three-dimensional batch data. The results showed that the monitoring performance was improved significantly in terms of detection success rate of process fault.

Keywords: Process Monitoring, kernel methods, multivariate filtering, data-driven techniques, quality improvement.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1713

Authors: Dinesh Rao, M.G.M. Khan, Sabiha Khan

Abstract:

Calibration estimation is a method of adjusting the original design weights to improve the survey estimates by using auxiliary information such as the known population total (or mean) of the auxiliary variables. A calibration estimator uses calibrated weights that are determined to minimize a given distance measure to the original design weights while satisfying a set of constraints related to the auxiliary information. In this paper, we propose a new multivariate calibration estimator for the population mean in the stratified sampling design, which incorporates information available for more than one auxiliary variable. The problem of determining the optimum calibrated weights is formulated as a Mathematical Programming Problem (MPP) that is solved using the Lagrange multiplier technique.

Keywords: Calibration estimation, Stratified sampling, Multivariate auxiliary information, Mathematical programming problem, Lagrange multiplier technique.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1899

Authors: Cristina G. Dascâlu, Corina Dima Cozma, Elena Carmen Cotrutz

Abstract:

The medical data statistical analysis often requires the using of some special techniques, because of the particularities of these data. The principal components analysis and the data clustering are two statistical methods for data mining very useful in the medical field, the first one as a method to decrease the number of studied parameters, and the second one as a method to analyze the connections between diagnosis and the data about the patient-s condition. In this paper we investigate the implications obtained from a specific data analysis technique: the data clustering preceded by a selection of the most relevant parameters, made using the principal components analysis. Our assumption was that, using the principal components analysis before data clustering - in order to select and to classify only the most relevant parameters – the accuracy of clustering is improved, but the practical results showed the opposite fact: the clustering accuracy decreases, with a percentage approximately equal with the percentage of information loss reported by the principal components analysis.

Keywords: Data clustering, medical data, principal components analysis.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1462

Authors: M. Giannopoulou, I. Roukounis, A. Roukouni.

Abstract:

Geographical Information Systems are an integral part of planning in modern technical systems. Nowadays referred to as Spatial Decision Support Systems, as they allow synergy database management systems and models within a single user interface machine and they are important tools in spatial design for evaluating policies and programs at all levels of administration. This work refers to the creation of a Geographical Information System in the context of a broader research in the area of influence of an under construction station of the new metro in the Greek city of Thessaloniki, which included statistical and multivariate data analysis and diagrammatic representation, mapping and interpretation of the results.

Keywords: Databases, Geographical information systems (GIS), Land-use planning, Metro stations

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1561

Authors: H. Izadbakhsh, M.Hour Ali, A. Amirkhani, A. Montazeri, M. Saberi

Abstract:

The after–sales activities are nowadays acknowledged as a relevant source of revenue, profit and competitive advantage in most manufacturing industries. Top and middle management, therefore, should focus on the definition of a structured business performance measurement system for the after-sales business. The paper aims at filling this gap, and presents an integrated methodology for the after-sales network performance measurement, and provides an empirical application to automotive case companies and their official service network. This is the first study that presents an integrated multivariate approach for total assessment and improvement of after-sale services.

Keywords: Data Envelopment Analysis (DEA), Principal Component Analysis (PCA), Automotive companies, After-sale services.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1858

Authors: Victor Breux, Jérôme Boutet, Alain Goret, Viviane Cattin

Abstract:

Early detection of anomalies in data centers is important to reduce downtimes and the costs of periodic maintenance. However, there is little research on this topic and even fewer on the fusion of sensor data for the detection of abnormal events. The goal of this paper is to propose a method for anomaly detection in data centers by combining sensor data (temperature, humidity, power) and deep learning models. The model described in the paper uses one autoencoder per sensor to reconstruct the inputs. The auto-encoders contain Long-Short Term Memory (LSTM) layers and are trained using the normal samples of the relevant sensors selected by correlation analysis. The difference signal between the input and its reconstruction is then used to classify the samples using feature extraction and a random forest classifier. The data measured by the sensors of a data center between January 2019 and May 2020 are used to train the model, while the data between June 2020 and May 2021 are used to assess it. Performances of the model are assessed a posteriori through F1-score by comparing detected anomalies with the data center’s history. The proposed model outperforms the state-of-the-art reconstruction method, which uses only one autoencoder taking multivariate sequences and detects an anomaly with a threshold on the reconstruction error, with an F1-score of 83.60% compared to 24.16%.

Keywords: Anomaly detection, autoencoder, data centers, deep learning.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 680

Authors: Jaime R. S. Fonseca

Abstract:

Nevertheless the widespread application of finite mixture models in segmentation, finite mixture model selection is still an important issue. In fact, the selection of an adequate number of segments is a key issue in deriving latent segments structures and it is desirable that the selection criteria used for this end are effective. In order to select among several information criteria, which may support the selection of the correct number of segments we conduct a simulation study. In particular, this study is intended to determine which information criteria are more appropriate for mixture model selection when considering data sets with only categorical segmentation base variables. The generation of mixtures of multinomial data supports the proposed analysis. As a result, we establish a relationship between the level of measurement of segmentation variables and some (eleven) information criteria-s performance. The criterion AIC3 shows better performance (it indicates the correct number of the simulated segments- structure more often) when referring to mixtures of multinomial segmentation base variables.

Keywords: Quantitative Methods, Multivariate Data Analysis, Clustering, Finite Mixture Models, Information Theoretical Criteria, Simulation experiments.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1484

Authors: Azadeh Golshan, Craig Evans, Phillip Geary, Abigail Morrow, Zoe Rogers, Marcel Maeder

Abstract:

22 physicochemical variables have been determined in water samples collected weekly from January to December in 2013 from three sampling stations located within a major drinking water reservoir. Classical Multivariate Curve Resolution Alternating Least Squares (MCR-ALS) analysis was used to investigate the environmental factors associated with the physico-chemical variability of the water samples at each of the sampling stations. Matrix augmentation MCR-ALS (MA-MCR-ALS) was also applied, and the two sets of results were compared for interpretative clarity. Links between these factors, reservoir inflows and catchment land-uses were investigated and interpreted in relation to chemical composition of the water and their resolved geographical distribution profiles. The results suggested that the major factors affecting reservoir water quality were those associated with agricultural runoff, with evidence of influence on algal photosynthesis within the water column. Water quality variability within the reservoir was also found to be strongly linked to physical parameters such as water temperature and the occurrence of thermal stratification. The two methods applied (MCR-ALS and MA-MCR-ALS) led to similar conclusions; however, MA-MCR-ALS appeared to provide results more amenable to interpretation of temporal and geological variation than those obtained through classical MCR-ALS.

Keywords: Catchment management, drinking water reservoir, multivariate curve resolution alternating least squares, thermal stratification, water quality.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 878

Authors: Lily Ingsrisawang, Tasanee Nacharoen

Abstract:

The problems arising from unbalanced data sets generally appear in real world applications. Due to unequal class distribution, many researchers have found that the performance of existing classifiers tends to be biased towards the majority class. The k-nearest neighbors’ nonparametric discriminant analysis is a method that was proposed for classifying unbalanced classes with good performance. In this study, the methods of discriminant analysis are of interest in investigating misclassification error rates for classimbalanced data of three diabetes risk groups. The purpose of this study was to compare the classification performance between parametric discriminant analysis and nonparametric discriminant analysis in a three-class classification of class-imbalanced data of diabetes risk groups. Data from a project maintaining healthy conditions for 599 employees of a government hospital in Bangkok were obtained for the classification problem. The employees were divided into three diabetes risk groups: non-risk (90%), risk (5%), and diabetic (5%). The original data including the variables of diabetes risk group, age, gender, blood glucose, and BMI were analyzed and bootstrapped for 50 and 100 samples, 599 observations per sample, for additional estimation of the misclassification error rate. Each data set was explored for the departure of multivariate normality and the equality of covariance matrices of the three risk groups. Both the original data and the bootstrap samples showed nonnormality and unequal covariance matrices. The parametric linear discriminant function, quadratic discriminant function, and the nonparametric k-nearest neighbors’ discriminant function were performed over 50 and 100 bootstrap samples and applied to the original data. Searching the optimal classification rule, the choices of prior probabilities were set up for both equal proportions (0.33: 0.33: 0.33) and unequal proportions of (0.90:0.05:0.05), (0.80: 0.10: 0.10) and (0.70, 0.15, 0.15). The results from 50 and 100 bootstrap samples indicated that the k-nearest neighbors approach when k=3 or k=4 and the defined prior probabilities of non-risk: risk: diabetic as 0.90: 0.05:0.05 or 0.80:0.10:0.10 gave the smallest error rate of misclassification. The k-nearest neighbors approach would be suggested for classifying a three-class-imbalanced data of diabetes risk groups.

Keywords: Bootstrap, diabetes risk groups, error rate, k-nearest neighbors.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1974

Authors: Emmanuel U. Ohaegbulem, Felix N. Nwobi

Abstract:

Alternative and simple computational arrangements in carrying out multivariate profile analysis when more than two groups (populations) are involved are presented. These arrangements have been demonstrated to not only yield equivalent results for the test statistics (the Wilks lambdas), but they have less computational efforts relative to other arrangements so far presented in the literature; in addition to being quite simple and easy to apply.

Keywords: Coincident profiles, g-group profile analysis, level profiles, parallel profiles, repeated measures MANOVA.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1203