Univariate

14,000,000 Leading Edge Experts on the ideXlab platform

Scan Science and Technology

Contact Leading Edge Experts & Companies

Scan Science and Technology

Contact Leading Edge Experts & Companies

The Experts below are selected from a list of 496071 Experts worldwide ranked by ideXlab platform

Olcay Taner Yildiz - One of the best experts on this subject based on the ideXlab platform.

  • VC-Dimension of Univariate Decision Trees
    IEEE transactions on neural networks and learning systems, 2015
    Co-Authors: Olcay Taner Yildiz
    Abstract:

    In this paper, we give and prove the lower bounds of the Vapnik–Chervonenkis (VC)-dimension of the Univariate decision tree hypothesis class. The VC-dimension of the Univariate decision tree depends on the VC-dimension values of its subtrees and the number of inputs. Via a search algorithm that calculates the VC-dimension of Univariate decision trees exhaustively, we show that our VC-dimension bounds are tight for simple trees. To verify that the VC-dimension bounds are useful, we also use them to get VC-generalization bounds for complexity control using structural risk minimization in decision trees, i.e., pruning. Our simulation results show that structural risk minimization pruning using the VC-dimension bounds finds trees that are more accurate as those pruned using cross validation.

  • Parallel Univariate decision trees
    Pattern Recognition Letters, 2007
    Co-Authors: Olcay Taner Yildiz, Onur Dikmen
    Abstract:

    Univariate decision tree algorithms are widely used in data mining because (i) they are easy to learn (ii) when trained they can be expressed in rule based manner. In several applications mainly including data mining, the dataset to be learned is very large. In those cases it is highly desirable to construct Univariate decision trees in reasonable time. This may be accomplished by parallelizing Univariate decision tree algorithms. In this paper, we first present two different Univariate decision tree algorithms C4.5 and Univariate linear discriminant tree. We show how to parallelize these algorithms in three ways: (i) feature based; (ii) node based; (iii) data based manners. Experimental results show that performance of the parallelizations highly depend on the dataset and the node based parallelization demonstrate good speedups.

Ma Chunsheng - One of the best experts on this subject based on the ideXlab platform.

Onur Dikmen - One of the best experts on this subject based on the ideXlab platform.

  • Parallel Univariate decision trees
    Pattern Recognition Letters, 2007
    Co-Authors: Olcay Taner Yildiz, Onur Dikmen
    Abstract:

    Univariate decision tree algorithms are widely used in data mining because (i) they are easy to learn (ii) when trained they can be expressed in rule based manner. In several applications mainly including data mining, the dataset to be learned is very large. In those cases it is highly desirable to construct Univariate decision trees in reasonable time. This may be accomplished by parallelizing Univariate decision tree algorithms. In this paper, we first present two different Univariate decision tree algorithms C4.5 and Univariate linear discriminant tree. We show how to parallelize these algorithms in three ways: (i) feature based; (ii) node based; (iii) data based manners. Experimental results show that performance of the parallelizations highly depend on the dataset and the node based parallelization demonstrate good speedups.

Thomas W. Yee - One of the best experts on this subject based on the ideXlab platform.

  • Univariate Continuous Distributions
    Springer Series in Statistics, 2015
    Co-Authors: Thomas W. Yee
    Abstract:

    This chapter enumerates those Univariate continuous distributions currently represented as VGLMs/VGAMs and implemented in VGAM. Most are grouped and tabulated according to their support, and/or the distribution from which they are derived (e.g., beta-type, gamma-type), and/or their purpose (e.g., statistical size distributions, actuarial distributions).

  • Univariate Discrete Distributions
    Springer Series in Statistics, 2015
    Co-Authors: Thomas W. Yee
    Abstract:

    This chapter and the next enumerates over 70 Univariate discrete and continuous distributions as VGLMs/VGAMs which are currently implemented in VGAM. Other variants, such as positive (zero-truncated), zero-inflated and zero-altered models, are described in a later chapter. A section is devoted to negative binomial regression, and it is shown that the VGLM/VGAM framework allows quite a number of variants to be naturally fitted.

Frank Noack - One of the best experts on this subject based on the ideXlab platform.

  • Prognostic Impact of VEGF and VEGF Receptor 1 (FLT1) Expression in Patients Irradiated for Stage II/III Non-Small Cell Lung Cancer (NSCLC)
    Strahlentherapie und Onkologie, 2010
    Co-Authors: Dirk Rades, Cornelia Setter, Juergen Dunst, Olav Dahl, Steven E. Schild, Frank Noack
    Abstract:

    Background and Purpose: The prognostic value of the tumor expression of vascular endothelial growth factor (VEGF) and VEGF receptor 1 (FLT1) is still unclear. This study investigated the impact of tumor expression of VEGF and FLT1 on outcomes in 61 patients irradiated for stage II/III non-small cell lung cancer (NSCLC). Material and Methods: The impact of tumor VEGF and FLT expression and twelve additional potential prognostic factors on locoregional control (LC), metastases-free survival (MFS), and overall survival (OS) were retrospectively evaluated. These factors included age, gender, performance status, histology, histological grade, T-category, N-category, surgery, chemotherapy, pack-years, smoking during radiotherapy, and hemoglobin levels during radiotherapy. Results: On Univariate analysis, improved LC was associated with lower T-category (p = 0.046), lower N-category (p = 0.047), and not smoking during radiotherapy (p = 0.012). VEGF (p = 0.26) and FLT1 expression (p = 0.70) had no significant impact. On multivariate analysis, lower N-category (p = 0.037) maintained significance; not smoking during radiotherapy was almost significant (p = 0.052). On Univariate analysis, improved MFS was associated with lower T-category (p = 0.034) and lower N-category (p = 0.027), and almost with hemoglobin ≥ 12 g/dl during radiotherapy (p = 0.053). VEGF (p = 0.80) and FLT1 expression (p = 0.61) had no significant impact. On multivariate analysis, lower N-category (p = 0.040) maintained significance. On Univariate analysis, improved OS was associated with lower T-category (p = 0.028), lower N-category (p = 0.003), not smoking during radiotherapy (p = 0.047), and hemoglobin levels ≥ 12 g/dl during radiotherapy (p = 0.019). VEGF (p = 0.59) and FLT1 expression (p = 0.85) had no significant impact. On multivariate analysis, lower N-category (p = 0.011) maintained significance. Conclusion: Tumor expression of VEGF and FLT1 appear to have no significant impact on LC, MFS, or OS in patients irradiated for NSCLC. Hintergrund und Ziel: Die prognostische Bedeutung der Tumorexpression von vaskularem endothelialen Wachstumsfaktor (VEGF) und VEGF-Rezeptor 1 (FLT1) ist unklar. Diese Studie untersuchte den Einfluss der Tumorexpression von VEGF und FLT1 auf die Prognose nach Strahlentherapie von 61 Patienten mit einem nichtkleinzelligen Bronchialkarzinom (NSCLC) im Stadium II/III. Material und Methodik: Der Einfluss von VEGF und FLT sowie zwolf weiteren möglichen Prognosefaktoren auf die lokoregionale Kontrolle (LC), das metastasenfreie Überleben (MFS) und das Gesamtüberleben (OS) wurde retrospektiv evaluiert. Diese Faktoren waren Alter, Geschlecht, Allgemeinzustand, Histologie, histologisches Grading, T-Kategorie, N-Kategorie, Operation, Chemotherapie, Pack-Years, Rauchen während Strahlentherapie und Hamoglobinwerte während Strahlentherapie (Tabelle 1). Ergebnisse: In der Univariaten Analyse war eine bessere LC mit niedrigerer T-Kategorie (p = 0,046), niedrigerer N-Kategorie (p = 0,047) und Nichtrauchen während Strahlentherapie (p = 0,012) assoziiert (Tabelle 2). VEGF- (p = 0,26) und FLT1-Expression (p = 0,70) waren nicht signifikant (Abbildung 1). In der Multivarianzanalyse blieb die N-Kategorie (p = 0,037) signifikant; Nichtrauchen war fast signifikant (p = 0,052). In der Univariaten Analyse war ein besseres MFS mit niedrigerer T-Kategorie (p = 0,034), niedrigerer N-Kategorie (p = 0,027) und fast mit Hamoglobinwerten ≥ 12 g/dl (p = 0,053) assoziiert (Tabelle 3). VEGF- (p = 0,80) und FLT1-Expression (p = 0,61) waren nicht signifikant (Abbildung 2). In der Multivarianzanalyse blieb die N-Kategorie (p = 0,040) signifikant. In der Univariaten Analyse war ein besseres OS mit niedrigerer T-Kategorie (p = 0,028), niedrigerer N-Kategorie (p = 0,003), Nichtrauchen (p = 0,047) und Hamoglobinwerten ≥ 12 g/dl (p = 0,019) assoziiert (Tabelle 4). VEGF- (p = 0,59) und FLT1-Expression (p = 0,85) waren nicht signifikant (Abbildung 3). In der Multivarianzanalyse blieb die N-Kategorie (p = 0,011) signifikant. Schlussfolgerung: Die Tumorexpression von VEGF und FLT1 scheint keinen signifikanten Einfluss auf die Prognose nach Bestrahlung aufgrund eines NSCLC im Stadium II/III zu haben.