Multiple-Instance Classification via Generalized Eigenvalue Proximal SVM

2010 ◽  
Vol 143-144 ◽  
pp. 1235-1239
Author(s):  
Zhen Wang ◽  
Dong Mei Li

The multiple-instance classification problem is formulated using a linear or nonlinear kernel as the minimization of a linear function in a finite dimensional real space subject to linear and bilinear constraints by SVM-based methods. This paper presents a new multiple-instance classifier that determines two nonparallel planes by solving generalized eigenvalue proximal SVM. Our method converges in a few iterations to a local solution. Computational results on a number of datasets indicate that the proposed algorithm is competitive with the other SVM-based methods in multiple-instance classification.

Mathematics ◽  
2021 ◽  
Vol 9 (18) ◽  
pp. 2346
Author(s):  
Almudena Campos-Jiménez ◽  
Francisco Javier García-Pacheco

In this paper we provide new geometric invariants of surjective isometries between unit spheres of Banach spaces. Let X,Y be Banach spaces and let T:SX→SY be a surjective isometry. The most relevant geometric invariants under surjective isometries such as T are known to be the starlike sets, the maximal faces of the unit ball, and the antipodal points (in the finite-dimensional case). Here, new geometric invariants are found, such as almost flat sets, flat sets, starlike compatible sets, and starlike generated sets. Also, in this work, it is proved that if F is a maximal face of the unit ball containing inner points, then T(−F)=−T(F). We also show that if [x,y] is a non-trivial segment contained in the unit sphere such that T([x,y]) is convex, then T is affine on [x,y]. As a consequence, T is affine on every segment that is a maximal face. On the other hand, we introduce a new geometric property called property P, which states that every face of the unit ball is the intersection of all maximal faces containing it. This property has turned out to be, in a implicit way, a very useful tool to show that many Banach spaces enjoy the Mazur-Ulam property. Following this line, in this manuscript it is proved that every reflexive or separable Banach space with dimension greater than or equal to 2 can be equivalently renormed to fail property P.


1989 ◽  
Vol 67 (9) ◽  
pp. 896-903 ◽  
Author(s):  
Lorenzo Resca

We show that a one-dimensional analytical study allows us to test and clarify the derivation, assumptions, and symmetry properties of the intervalley effective mass equation (IVEME). In particular, we show that the IVEME is consistent with a two-band case, and is in fact exact for a model that satisfies exactly all its assumptions. On the other hand, an alternative formulation in k-space that includes intervalley kinetic energy terms is consistent with a one-band case, provided that intra-valley kinetic energy terms are also calculated consistent with one band. We also show that the standard symmetry assumptions for both real space and k-space formulations are not actually exact, but are consistent with a "total symmetric" projection, or with taking spherical averages in a three-dimensional case.


1984 ◽  
Vol 62 (1) ◽  
pp. 171-177 ◽  
Author(s):  
Rizwan Ahmad ◽  
David A. Armstrong

Elementary considerations indicate that thiol radicals, RS•, should have a high redox potential [Formula: see text][Formula: see text]However, the equilibrium [4],[Formula: see text]which is established in the presence of excess RS−, would convert RS•to [Formula: see text] which is a reducing species. Experimentally it was demonstrated that thiol radicals made by γ radiolysis of β-mercaptoethanol solutions effected two-electron oxidation of dihydroflavin FlH2 at pH 6.3 and of FlH− at pH 8. On the other hand, [Formula: see text] readily reduced Fl to FlH2 or FlH− as expected. At pH 9, photostationary states were established after a few minutes radiolysis and the ratios [FlH−]ss/[Fl]ss were a function of [Formula: see text] The main reactions occurring were:[Formula: see text]The values of k19 and k22 were both large. The ratio k19/k22 was ∼0.8 for lumiflavin and ∼0.3 for flavin adenine dinucleotide. The cyclic disulphide anions of lipoamide and dithiothreitol [Formula: see text] also effected two-electron reductions of flavins. However, the protonated form of [Formula: see text] oxidized FlH2, and the photostationary ratio [FlH−]ss/[Fl]ss was an approximate linear function of [Formula: see text]. The implications of the observed changes in redox properties of sulphur radicals on complexation with RS− and protonation were briefly considered.Des considérations élémentaires indiquent que les radicaux thiyles, RS•, doivent avoir un potentiel rédox élevé [Formula: see text][Formula: see text]


Author(s):  
Masashi Okubo ◽  
Tomio Watanabe

In this paper, a collaboration support system for 3D shape evaluation by using network virtual reality is proposed on the basis of both studies in 3D shape evaluation and communication support in virtual space. Collaboration mainly consists of two tasks; one is each partner’s task that is performed by the individual, the other is communication with each other. Both of them are very important objectives for all the collaboration support system. The proposed system provides the two viewpoints for each task. One is the viewpoint of back side of user’s own avatar for the smooth communication. The other is that of avatar’s eye for 3D shape evaluation. Switching the viewpoints satisfies the task conditions for 3D shape evaluation and communication. The system basically consists of PC, HMD and sensors, and users can share the embodied interaction by observing interaction between their avatars in virtual space. However, the HMD and magnetic sensors, which are put on the users, would restrict the nonverbal communication. Then, we have tried to compensate the loss of nodding of partner’s avatar by introducing the speech-driven embodied interactive actor InterActor. Sensory evaluation by paired comparison of 3D shapes in the collaborative situation in virtual space and in real space and the questionnaire are performed. The result demonstrates the effectiveness of InterActor’s nodding in the collaborative situation.


2016 ◽  
Vol 2016 ◽  
pp. 1-14 ◽  
Author(s):  
Youngchul Han ◽  
Geonwoo Kim

We propose an efficient lattice method for valuation of options with barrier in a regime switching model. Specifically, we extend the trinomial tree method of Yuen and Yang (2010) by calculating the local average of prices near a node of the lattice. The proposed method reduces oscillations of the lattice method for pricing barrier options and improves the convergence speed. Finally, computational results for the valuation of options with barrier show that the proposed method with interpolation is more efficient than the other tree methods.


2015 ◽  
Vol 26 (2) ◽  
pp. 261-287 ◽  
Author(s):  
Johanna Mesch ◽  
Eli Raanes ◽  
Lindsay Ferrara

AbstractThis article reports on a linguistic study examining the use of real space blending in the tactile signed languages of Norwegian and Swedish signers who are both deaf and blind. Tactile signed languages are typically produced by interactants in contact with each other’s hands while signing. Of particular interest to this study are utterances which not only consist of the signer producing signs with his or her own hands (or other body parts), but which also recruit the other interactant’s hands (or another body part). These utterances, although perhaps less frequent, are co-constructed, in a very real sense, and they illustrate meaning construction during emerging, embodied discourse. Here, we analyze several examples of these types of utterances from a cognitive linguistic and cognitive semiotic perspective to explore how interactants prompt meaning construction through touch and the involvement of each other’s bodies during a particular type of co-regulation.


1970 ◽  
Vol 116 (535) ◽  
pp. 651-655 ◽  
Author(s):  
A. E. Maxwell

In a typical classification problem each subject in a sample of N subjects is allocated to one or other of k exhaustive and mutually exclusive categories. For example, a sample of families may be classified into social-class groupings in accordance with the Registrar General's classification based on the occupation of the father or father substitute. Or a sample of psychiatric patients may be classified into one or other of the diagnostic categories recommended by the W.H.O. Now it occasionally happens that the same sample of subjects is classified, for a given set of categories, independently by two different agents. For example in the Plowden Reports, Appendix 3 Table 29, a sample of children is classified into types of secondary school, Grammar, Comprehensive, Technical, etc., on the one hand according to the parent's ambition for their children and on the other hand according to the type of school in which the children were eventually placed. In situations such as the latter a k k classification table evolves in which, when the categories are arranged in the same order, the matches between the two separate classifications appear in the cells of the main diagonal of the table and the mismatches appear in the off-diagonal cells (see Table I below). The question then arises as to how to compare the two separate classifications and to measure in quantitative terms the degree of agreement between them.


2021 ◽  
Vol 24 (68) ◽  
pp. 123-137
Author(s):  
Sami Nasser Lauar ◽  
Mario Mestria

In this work, we present a metaheuristic based on the genetic and greedy algorithms to solve an application of the set covering problem (SCP), the data aggregator positioning in smart grids. The GGH (Greedy Genetic Hybrid) is structured as a genetic algorithm, but it has many modifications compared to the classic version. At the mutation step, only columns included in the solution can suffer mutation and be removed. At the recombination step, only columns from the parent’s solutions are available to generate the offspring. Moreover, the greedy algorithm generates the initial population, reconstructs solutions after mutation, and generates new solutions from the recombination step. Computational results using OR-Library problems showed that the GGH reached optimal solutions for 40 instances in a total of 75 and, in the other instances, obtained good and promising values, presenting a medium gap of 1,761%.


Author(s):  
D. Attaf ◽  
K. Djerriri ◽  
D. Mansour ◽  
D. Hamdadou

<p><strong>Abstract.</strong> Mapping of burned areas caused by forest fires was always a main concern to researchers in the field of remote sensing. Thus, various spectral indices and classification techniques have been proposed in the literature. In such a problem, only one specific class is of real interest and could be referred to as a one-class classification problem. One-class classification methods are highly desirable for quick mapping of classes of interest. A common used solution to deal with One-Class classification problem is based on oneclass support vector machine (OC-SVM). This method has proved useful in classification of remote sensing images. However, overfitting problem and difficulty in tuning parameters have become the major obstacles for this method. The new Presence and Background Learning (PBL) framework does not require complicated model selection and can generate very high accuracy results. On the other hand the Google Earth Engine (GEE) portal provides access to satellite and other ancillary data, cloud computing, and algorithms for processing large amounts of data with relative ease. Therefore, this study mainly aims to investigate the possibility of using the PBL framework within the GEE platform to extract burned areas from freely available Landsat archive in the year 2015. The quality of the results obtained using PBL framework was assessed using ground truth digitized by qualified technicians and compared to other classification techniques: Thresholding burned area spectral Index (BAI) and OC-SVM classifiers. Experimental results demonstrate that PBL framework for mapping the burned areas shows the higher classification accuracy than the other classifiers, and it highlights the suitability for the cases with few positive labelled samples available, which facilitates the tedious work of manual digitizing.</p>


2021 ◽  
Vol 16 (1) ◽  
pp. 12
Author(s):  
Biko Nabih Fikri Zufar ◽  
Eka Kartika Sari

Technological developments in society are separate space into two forms, real space and virtual space. The two meet each other when either individuals or communities use virtual space. Foucault described space to be three forms: dystopia, utopia, and heterotopia. This article aims to explain Instagram as virtual space that intersects real space, forming other spaces in the process. This article uses the perspective of the sociology of space and place while also using Foucault’s theory of other space. This study used the qualitative research method with a discourse analysis approach. The finding of this article is that there are six heterotopia principles contained in Instagram’s features. The results, Instagram forms other space beyond the activity of uploading photos or videos. Instagram as a real space becomes biased because of heterotopia. Individuals interpret Instagram differently as a place to share. On the contrary, when sharing activities, other individuals may use it for more than just sharing. Instagram has changed its space due to a shift in place use. Instagram can create other spaces besides its original function of uploading photos or videos through the other features on Instagram.


Sign in / Sign up

Export Citation Format

Share Document