Process analysis,monitoring and diagnosis,using multivariate projection methods

Multivariate statistical methods for the analysis, monitoring and diagnosis of process operating performance are becoming more important because of the availability of on-line process computers which routinely collect measurements on large numbers of process variables. Traditional univariate control charts have been extended to multivariate quality control situations using the Hotelling T² statistic. Recent approaches to multivariate statistical process control which utilize not only product quality data (Y), but also all of the available process variable data (X) are based on multivariate statistical projection methods (principal component analysis, (PCA), partial least squares, (PLS), multi-block PLS and multi-way PCA). An overview of these methods and their use in the statistical process control of multivariate continuous and batch processes is presented. Applications are provided on the analysis of historical data from the catalytic cracking section of a large petroleum refinery, on the monitoring and diagnosis of a continuous polymerization process and on the monitoring of an industrial batch process.     

Scientometric analysis of a scientific school

A procedure including scientometric methods combined with other techniques is described. This allows to define the membership and structure of a scientific school and to trace the dynamics of its development. Formation and evolution of a scientific school is presented as a purposeful scientific-information process involving changes in the nature and dynamics of the types of information links.     

A geostatistical analysis of geostatistics

The bibliometric indices of the scientific field of geostatistics were analyzed using statistical and spatial data analysis. The publications and their citation statistics were obtained from the Web of Science (4000 most relevant), Scopus (2000 most relevant) and Google Scholar (5389). The focus was on the analysis of the citation rate (CR), i.e. number of citations an author or a library item receives on average per year. This was the main criterion used to analyze global trends in geostatistics and to extract the Top 25 most-cited lists of the research articles and books in geostatistics. It was discovered that the average citation rate for geostatisticians has stabilized since 1999, while the authors’ n-index seems to have declined ever since. One reason for this may be because there are more and more young authors with a lower n-index. We also found that the number of publications an author publishes explains only 60% of the variation in the citation statistics and that this number progressively declines for an author with a lower number of publications. Once the geographic location is attached to a selection of articles, an isotropic Gaussian kernel smoother weighted by the CR can be used to map scientific excellence around the world. This revealed clusters of scientific excellence around locations such as Wageningen, London, Utrecht, Hampshire, UK, Norwich, Paris, Louvain, Barcelona, and Zürich (Europe); Stanford, Ann Arbor, Tucson, Corvallis, Seattle, Boulder, Montreal, Baltimore, Durham, Santa Barbara and Los Angeles (North America); and Canberra, Melbourne, Sydney, Santiago (Chile), Taipei, and Beijing (other continents). Further correlation with socio-economic variables showed that the spatial distribution of CRs in geostatistics is independent of the night light image (which represents economic activity) and population density. This study demonstrates that the commercial scientific indexing companies could enhance their service by assigning the geographical location to library items to allow spatial exploration and analysis of bibliometric indices.     

Scientometric analysis of nature, the journal

300,000 reports inNature during the 1869–1998 period have been reviewed. The distribution of articles by subfields was determined. Additional sources of information were several journals on analytical chemistry and papers at the Pittsburg conference series during 1950–1999. The methodology used is based on the analysis of the average age of employed instruments. The agreement between scientometric data from various sources of information depends on the development stage of the field of science. Calculated and measured scientometric curves were compared. One of the key trends in the development of basic sciences, namely, the increase of articles dealing with instrumental analytical chemistry, inNature is revealed. The research was supported by the Russian Foundation of Basic Research (Project No. 97-06-80163a) and State Target Programme "Integracia" (Project No. 670).     

Scientometric analysis of the major Iranian medical universities

Nowadays, scientometrics has become an important field of study to monitor the progresses in scientific performance of a research group, a department, a university etc. A number of scientometrical studies have been done about Iranian scientific outcome in recent years. But there is no comparison between major Iranian medical universities. In this study, by using Scopus as search engine, the scientific outcomes of the Iran University of Medical Sciences, Isfahan University of Medical Sciences, Mashhad University of Medical Sciences, Shahid Beheshti University of Medical Sciences, Shiraz University of Medical Sciences, Tabriz University of Medical Sciences, and Tehran University of Medical Sciences have been compared with each other. These universities were compared by the number of published articles per year, number of citations received per year, number of citations received per year per article, total H-indices, top ten authors, and top ten journals. The results of this study show that the order of the studied universities in research performance is as follow: Tehran > Shiraz = Shahid Beheshti > Isfahan = Iran > Tabriz = Mashhad universities of medical sciences. In addition, the data of Tehran University of Medical Sciences as the top medical university of Iran was compared with some of top medical universities around the world.     

Characterization of methanol-water and acetonitrile-water association using multivariate curve resolution methods

The solvatochromic comparison method has been used to probe the interactions of solutes with binary solvent mixtures of methanol-water and acetonitrile-water. The solute spectra recorded in these mixtures are composed of the additive spectral contributions of the different solvated species of the solute, i.e., the water-solvated species, the cosolvent-solvated species, and the species solvated by water-solvent complexes. Multivariate curve resolution-alternating least squares has been used to model the solvation of the solutes as a function of the composition of the binary solvent mixture. Spectra and concentration profiles of the dye surrounded by the different solvation environments have been isolated. For the first time, solute spectra solvated exclusively by methanol-water and acetonitrile-water complexes have been obtained, and the solvatochromic parameters of dipolarity/polarizability and hydrogen-bonding acidity have been estimated for these complex species.     

Scientometric analysis of attendance at International Scientific Meetings

International scientific meetings represent important channels for communicating research results. Based on data from more than 500 proceedings of scientific meetings, organization and participation patterns of several countries (or geopolitical regions) were analyzed. Some new indicators were derived and proved to be useful in characterizing the scientific activity of the countries. Particularly, the “open” and “closed” nature of national scientific communities, as well as “attraction” and “repulsion” between certain pairs of countries could be revealed by this method.     

Scientometric indicators: peer-review,bibliometric methods and conflict of interests

The paper discusses the role of scientometric indicators in peer-review selection of research project proposals. An ex post facto evaluation was made of three calls for research project proposals in Slovenia: 2003 with a peer review system designed in a way that conflict of interest was not avoided effectively, 2005 with a sound international peer-review system with minimized conflict of interest influence but a limited number of reviewers, and 2008 with a combination of scientometric indicators and a sound international peer review with minimized conflict of interest influence. The hypothesis was that the three different peer review systems would have different correlations with the same set of scientometric indicators. In the last two decision-making systems (2005 and 2008) where conflict of interest was effectively avoided, we have a high percentage (65%) of projects that would have been selected in the call irrespective of the method (peer review or bibliometrics solely). In contrast, in the 2003 call there is a significantly smaller percentage (49%) of projects that would have been selected in the call irrespective of the method (peer review or bibliometrics solely). It was shown that while scientometric indicators can hardly replace the peer-review system as the ultimate decision-making and support system, they can reveal its weaknesses on one hand and on the other can verify peer-review scores and minimize conflict of interest if necessary.     

Quantification of ATOFMS data by multivariate methods

Aerosol time-of-flight mass spectrometry (ATOFMS) is capable of measuring the sizes and chemical compositions of individual polydisperse aerosol particles in real time. A qualitative estimate of the particle composition is acquired in the form of a mass spectrum that must be subsequently interpreted in order to draw conclusions regarding atmospheric relevance. The actual problem involves developing a calibration that allows the mass spectral data to be transformed into estimates of the composition of the atmospheric aerosol. A properly calibrated ATOFMS system should be able to quantitatively determine atmospheric concentrations of various species. Ideally, it would be able to accomplish this more rapidly, accurately, with higher size and time resolution, and at a far lower marginal cost than the manual sampling methods that are currently employed. Attempts have already been made at using ATOFMS and similar techniques to extract the bulk chemical species concentration present in an ensemble of particles. This study represents the use of a multivariate calibration method, two-dimensional partial least-squares analysis, for calibrating single-particle mass spectral data. The method presented here is far less labor-intensive than the univariate methods attempted to date and allows for less observer bias. Because of the labor savings, this is also the most comprehensive calibration performed to date, resulting in the quantification of 44 different chemical species.     

Scientometric studies on chemistry II: Aims and methods of producing new chemical substances

Chemistry, as today's most active science, has increased its substances exponentially during the past 200 years without saturation. To get more insight why and how chemists produce new substances, a content analysis of 300 communications to theAngewandte Chemie of the years 1980, 1990, and 1995 is carried out regarding aims and methods of preparative research. In the most productive field of organic chemistry production mainly occurs to improve abilities for further production, while the less productive field of inorganic chemistry has more diverse aims. Methodological differences between organic and inorganic chemistry are discussed in detail as well as the relationship between pure and applied science.     

Scientometric analysis of relative performance in a key university in Romania

A scientometric analysis of the Babe?-Bolyai University in Romania is provided, highlighting the strong and the weak points with respect to a range of leading international universities and referencing to some extent to nation-wide data from several countries. Taken into account are such items as total number of publications, analyses per subject area or per research field, number of citations, types of publications, Hirsch indexes, and books. Internationally, chemistry, physics, mathematics, computer science, religion, area studies, geology, paleontology, and public administration are identified as the most active areas. Nationally, a number of additional strong points are identified, such as psychology, history, and environmental sciences. The percentage of researchers with reasonably high activity (e.g., at least ~ one publication per year as indexed in major databases) is relatively low (~10 %), and the percentage with reasonably high international competitiveness (based on citation counts, number of publications, books indexed in international libraries) is at only ~2 %. The decisive factor controlling an exponential increase in publications since ~2000–2004 appears to have been a conservatively managed exponential increase of the national GDP and implicitly of the research budgets.     

Principal alarms in multivariate statistical process control using independent component analysis

This article proposes a methodology that helps to predict the main mean shifts, denoted as principal alarms, in a non-normal multivariate process using the available in-control data. The analysis is based on the transformation of the observed correlated variables into independent factors using independent component analysis. These independent components allow us to simulate shifts preserving the covariance structure. The graphical representations of those simulated shifts are helpful in improving the design and control of the process. Two real manufacturing processes are presented showing the advantage of the proposed methodology.     

Scientometric analysis of the topical content of scientific research and its particularities

A new method for quantitative evaluation of the topical content of scientific research is proposed. The method is based on the analysis of the number and topic of publications in different fields. A mathematical model, describing the connection between level and width of research, between topic renewal and concentration of research is developed. Furthermore, coefficients characterizing various aspects of research are introduced. The theoretical conclusions fit well the factual data obtained from the INIS system. A nucleus of terms defining the most developed directions of research is found in each case. The analysis indicates that the growth rate of publications cannot serve as a reliable criterion of research topicality.     

A multivariate Poisson-lognormal regression model for prediction of crash counts by severity, using Bayesian methods

Numerous efforts have been devoted to investigating crash occurrence as related to roadway design features, environmental factors and traffic conditions. However, most of the research has relied on univariate count models; that is, traffic crash counts at different levels of severity are estimated separately, which may neglect shared information in unobserved error terms, reduce efficiency in parameter estimates, and lead to potential biases in sample databases. This paper offers a multivariate Poisson-lognormal (MVPLN) specification that simultaneously models crash counts by injury severity. The MVPLN specification allows for a more general correlation structure as well as overdispersion. This approach addresses several questions that are difficult to answer when estimating crash counts separately. Thanks to recent advances in crash modeling and Bayesian statistics, parameter estimation is done within the Bayesian paradigm, using a Gibbs Sampler and the Metropolis–Hastings (M–H) algorithms for crashes on Washington State rural two-lane highways. Estimation results from the MVPLN approach show statistically significant correlations between crash counts at different levels of injury severity. The non-zero diagonal elements suggest overdispersion in crash counts at all levels of severity. The results lend themselves to several recommendations for highway safety treatments and design policies. For example, wide lanes and shoulders are key for reducing crash frequencies, as are longer vertical curves.     

Non-linear analysis of wave propagation using transform methods

Non-linear wave propagation/transient dynamics in lattice structures is modeled using a technique which combines the Laplace transform and the Finite element method. The first step in the technique is to apply the Laplace transform to the governing differential equations and boundary conditions of the structural model. The non-linear terms present in these equations are represented in the transform domain by making use of the complex convolution theorem. Then, a weak formulation of the transformed equations yields a set of element level matrix equations. The trial and test functions used in the weak formulation correspond to the exact solutions of the linear parts of the transformed governing differential equations. Numerical results are presented for a viscoelastic rod and von Karman type beam.     

Quantitative Raman spectroscopy of highly fluorescent samples using pseudosecond derivatives and multivariate analysis

Intense luminescence backgrounds cause significant problems in quantitative Raman spectroscopy, particularly in multivariate analysis where background suppression is essential. Taking second derivatives reduces the background, but differentiation increases the apparent noise that arises on spectra recorded with CCD detectors due to random, but fixed, variations in the pixel-to-pixel response. We have recently reported a very general method for correcting CCD fixed-pattern response in which spectra are taken at two or more slightly shifted spectrometer positions and are then subtracted to give a derivative-like shifted, subtracted Raman (SSR) spectrum. Here we show that differentiating SSR data (which has inherently higher S/N than the undifferenced data) yields spectra that are similar to those that are obtained from the normal two-step differentiation process and can be characterized as pseudo-second-derivative, PSD, spectra. The backgrounds are suppressed in the PSD spectra, which means they can be used directly in multivariate data analysis, but they have significantly higher S/N ratios than do simple second derivatives. To demonstrate the improvement brought about by using PSD spectra, we have analyzed known samples, consisting of simple binary mixtures of methanol and ethanol doped with laser dye. When the background levels of all samples included in the models were < or =10x greater than the intensity of the strongest Raman bands, partial least-squares calibration of the PSD data gave a standard error of prediction of 3.2%. Calibration using second derivatives gave a prediction error which was approximately twice as large, at 6.5%; however, when data with background levels . approximately 100x larger than the strongest Raman bands were included, the noise on the second-derivative spectra was so large as to prevent a meaningful calibration. Conversely, the PSD treatment of these samples gave a very satisfactory calibration with a standard error of prediction (3.3%) almost identical to that obtained when the most fluorescent samples were excluded. This method clearly has great potential for general purpose Raman analytical chemistry, because it does not depend on specialized equipment, is computationally undemanding, and gives stable and robust calibrations, even for samples in which the luminescence background level fluctuates between the extremes of being practically zero and completely dominating the Raman signal.     

Scientometric study of laser patent literature

An analysis of the patents filed and scientific papers published and abstracted in theJournal of Current Laser Abstracts (JCLA) for the period 1967–95 indicates that innovative activity in laser science and technology was at its peak in the early 70s. However, scientific activity surpassed the innovative activity in the early 80s. There was a continuous shift in emphasis from “applications of lasers” to “experimental laser research” and to “theoretical laser research”. Further analysis of the 1840 patents field in 1970–71, 1975–76, and 1980–85 indicates that most of the firms filing patents were situated in USA and thus USA is the leading country filing patents in this area followed by Japan. “Spectroscopy of laser output” followed by “Communication applications of laser” got the maximum emphasis.