8 research outputs found
Operating characteristics for the design and optimisation of classification systems
No full textIn statistical pattern recognition, problems involve distinguishing of various concepts or classes, based on the development of classifiers/discriminators. These exploit discriminatory information existing in measurements originating from objects. A trained classifier results in a partitioning in measurement space, providing some separation between the various classes. In the (typical) case of class overlap, this partitioning inherently results in a trade-off between the various possible classification errors that may occur. This partitioning can be modified to adjust these trade-offs. Given class abundances, a classifier can be evaluated at a given partitioning. However, variations in the abundances leads to an altered classifier performance. These fundamental aspects behind classifier design and evaluation can be studied within the framework of classifier operating characteristics, which is the topic of this dissertation. The contents consist of a number of published/accepted journal and conference papers, contextualised into a number of chapters representing various aspects of operating characteristic analysis. First the well-known two-class operating characteristic is considered, with two new analyses that are useful in certain circumstances. Next, the extension to the elusive multiclass case is considered, showing how standard 2-class operating characteristics analyses can be extended theoretically to the multiclass case. The challenge behind the multiclass extension is shown to be of a computational nature, with the calculation size increasing exponentially with the number of classes. The primary thesis contribution is then presented, consisting of a number of approaches and philosophies that can be used to overcome the computational challenges. Of primary importance is the finding that most practical problems are such that not all dimensions of the operating characteristic interact together significantly. Next it is shown how the operating characteristic approach can be used to design classifiers in ill-defined environments. In these problems some classes may be poorly represented, and the goal of the classifier design is to protect against these unforeseen conditions. Finally, it is shown that operating characteristics can be applied to a multi-stage classifier setup, allowing for a holistic design incorporating interactions between classes, and the classifier stages.Electrical Engineering, Mathematics and Computer Scienc
Efficient Multiclass ROC Approximation by Decomposition via Confusion Matrix Perturbation Analysis
No full textReceiver operator characteristic (ROC) analysis has become a standard tool in the design and evaluation of two-class classification problems. It allows for an analysis that incorporates all possible priors, costs, and operating points, which is important in many real problems, where conditions are often nonideal. Extending this to the multiclass case is attractive, conferring the benefits of ROC analysis to a multitude of new problems. Even though theROCanalysis extends theoretically to the multiclass case, the exponential computational complexity as a function of the number of classes is restrictive. In this paper, we show that the multiclass ROC can often be simplified considerably because some ROC dimensions are independent of each other. We present an algorithm that analyzes interactions between various ROC dimensions, identifying independent classes, and groups of interacting classes, allowing the ROC to be decomposed. The resulting decomposed ROC hypersurface can be interrogated in a similar fashion to the ideal case, allowing for approaches such as cost sensitive and Neyman-Pearson optimization, as well as the volume under the ROC. An extensive bouquet of examples and experiments demonstrates the potential of this methodology.MediamaticsElectrical Engineering, Mathematics and Computer Scienc
Testing absolute plate reference frames and the implications for the generation of geodynamic mantle heterogeneity structure
No full textInternational audienceAbsolute reference frames are a means of describing the motion of plates on the surface of the Earth over time, relative to a fixed point or “frame.” Multiple models of absolute plate motion have been proposed for the Cretaceous–Tertiary period, however, estimating the robustness and limitations of each model remains a significant limitation for refining both regional and global models of plate motion as well as fully integrated and time dependent geodynamic models. Here, we use a novel approach to compare five models of absolute plate motion in terms of their consequences for forward modelled deep mantle structure since at least 140 Ma. We show that the use of hotspots, either fixed or moving, or palaeomagnetics, with or without corrections for true-polar wander, leads to significant differences in palaeo-plate velocities of over 10 cm/yr as well as differences in the location of palaeo-plate boundaries of up to 30° in longitude and latitude. Furthermore, we suggest that first order differences in forward predicted mantle structure between the models are due mostly to differences in palaeo-plate velocities, whereas variation in the location of plate boundaries may contribute to smaller wavelength offsets. We present a global comparison of the absolute reference frames in terms of mantle structure, which we have tomographically filtered to reflect the resolution of the seismic tomography model S20RTS. At very long wavelengths hotspot models best reproduce the mantle structure. However, when geometry and the match of smaller-scale subducted slab volumes are compared, a hybrid model based on moving hotspots after 100 Ma and palaeomagnetic data before (with no corrections for true-polar wander), best reproduces the overall mantle structure of slab burial grounds, even though no single model fits best at all mantle depths. We find also that the published subduction reference frame tested here results in a modelled mantle structure that agrees well with S20RTS for depths > 2500 km, equivalent to subduction before the Cretaceous, but not for shallower depths. This indicates that a careful assimilation of hotspot, palaeomagnetic and seismic tomography data into future absolute plate motion models is required to derive a more robust subduction reference frame
The European Trade in Stained Glass, with Special Reference to the Trade between the Rhineland and the United Kingdom 1794-1835
No full textThe study is set against a period of cultural and political change in Continental Europe and the United Kingdom at the beginning of the nineteenth century. As a result of the Concordat between Napoleon and Pope Pius VII signed on the 15 July 1801, the Pope agreed that he and his successors would take no action against the buyers of church property. In this way the secularising of church property, firstly in France and later in other countries, was legalised. The wholesale redistribution began during the Peace of Amiens in 1802, one of the “Goods” on offer being stained glass. The United Kingdom was the primary beneficiary of this redistribution with its emerging interest in historical objects, firstly antiquarian, then commercial and later intellectual.
Because of the extent of the trade in stained glass in this period the study concentrates on the trade between the Rhineland and the United Kingdom, with the focus being between 1815-1835, when a new constellation of buyers, sellers and installers of Rhenish stained glass appeared. The function of the appendices is twofold: firstly to remove detailed but relevant data from the main text so as not to obscure the main argument and secondly to provide the reader with data not strictly within the parameters of the thesis.
The analysis of the available and newly discovered data takes three approaches to acquisition and installation and is presented in three case studies. Firstly the activities of the Regency Contractor architects, highlighting Sir Jeffry Wyatville and William Wilkins who were responsible for the most significant Rhenish stained glass installations in this period. The second analyses Edward Spenser Curling (1771-1850) whose newly discovered diary (by the author) of his activities between 1827-36, sheds new light on the detailed mechanics of the stained glass trade of this period and the networks existing in Cologne and the United Kingdom. The discovery of this diary proves that Curling acquired all the Altenberg and St Apern stained glass panels presently in the United Kingdom. The third study analyses a number of facets that influenced the trade in the United Kingdom, particularly the activities of the stained glass painters Betton and Evans
Recommended from our members
Word Order, Focus, and Clause Linking in Greek Tragic Poetry
No full textThe thesis comprises an investigation of three aspects of sentence structure in Classical Greek (henceforth CG) dramatic poetry: order of the main sentence elements (subject, verb, and object) within the clause, the emphatic position at the start of the clause, and the structure of inter-clausal linking. It is argued that these three features, usually considered separately, are interdependent, and that intra-clausal word order is directly related to the structure of compound and complex sentences. The discussion undertakes a systematic survey of subject, verb, and object order in a corpus of texts, proposes an explanation for the observed order, and develops a model which explains how prominence within the clause is exploited in clause linking to produce the complement structures observed in Homeric and tragic complementation
Enhanced detection of early hepatocellular carcinoma by serum SELDI-TOF proteomic signature combined with alpha-fetoprotein marker
No full textBackground: Biomarkers for accurate diagnosis of early hepatocellular carcinoma (HCC) are limited in number and clinical validation. We applied SELDI-TOF-MS ProteinChip technology to identify serum profile for distinguishing HCC and liver cirrhosis (LC) and to compare the accuracy of SELDI-TOF-MS profile and alpha-fetoprotein (AFP) level in HCC diagnosis. Patients and Methods: Serum samples were obtained from 120 HCC and 120 LC patients for biomarker discovery and validation studies. ProteinChip technology was employed for generating SELDI-TOF proteomic features and analyzing serum proteins/peptides. Results: A diagnostic model was established by CART algorithm, which is based on 5 proteomic peaks with m/z values at 3324, 3994, 4665, 4795, and 5152. In the training set, the CART algorithm could differentiate HCC from LC subjects with a sensitivity and specificity of 98% and 95%, respectively. The results were assessed in blind validation using separate cohorts of 60 HCC and 60 LC patients, with an accuracy of 83% for HCC and 92% for LC patients. The diagnostic odd ratio (DOR) indicated that SELDI-TOF proteomic signature could achieve better diagnostic performance than serum AFP level at a cutoff of 20 ng/mL (AFP 20) (92.72 vs 9.11), particularly superior for early-stage HCC (87% vs 54%). Importantly, a combined use of both tests could enhance the detection of HCC (sensitivity, 95%; specificity, 98%; DOR, 931). Conclusion: Serum SELDI-TOF proteomic signature, alone or in combination with AFP marker, promises to be a good tool for early diagnosis and/screening of HCC in at-risk population with liver cirrhosis. © 2010 Society of Surgical Oncology.published_or_final_versionSpringer Open Choice, 01 Dec 201
