Home About us Contact
|
Home About us Contact | ||
|
|
||
Tree Algorithm (tree + algorithm)
Selected AbstractsInducing safer oblique trees without costsEXPERT SYSTEMS, Issue 4 2005Sunil Vadera Abstract: Decision tree induction has been widely studied and applied. In safety applications, such as determining whether a chemical process is safe or whether a person has a medical condition, the cost of misclassification in one of the classes is significantly higher than in the other class. Several authors have tackled this problem by developing cost-sensitive decision tree learning algorithms or have suggested ways of changing the distribution of training examples to bias the decision tree learning process so as to take account of costs. A prerequisite for applying such algorithms is the availability of costs of misclassification. Although this may be possible for some applications, obtaining reasonable estimates of costs of misclassification is not easy in the area of safety. This paper presents a new algorithm for applications where the cost of misclassifications cannot be quantified, although the cost of misclassification in one class is known to be significantly higher than in another class. The algorithm utilizes linear discriminant analysis to identify oblique relationships between continuous attributes and then carries out an appropriate modification to ensure that the resulting tree errs on the side of safety. The algorithm is evaluated with respect to one of the best known cost-sensitive algorithms (ICET), a well-known oblique decision tree algorithm (OC1) and an algorithm that utilizes robust linear programming. [source] A sparse marker extension tree algorithm for selecting the best set of haplotype tagging single nucleotide polymorphismsGENETIC EPIDEMIOLOGY, Issue 4 2005Ke Hao Abstract Single nucleotide polymorphisms (SNPs) play a central role in the identification of susceptibility genes for common diseases. Recent empirical studies on human genome have revealed block-like structures, and each block contains a set of haplotype tagging SNPs (htSNPs) that capture a large fraction of the haplotype diversity. Herein, we present an innovative sparse marker extension tree (SMET) algorithm to select optimal htSNP set(s). SMET reduces the search space considerably (compared to full enumeration strategy), and therefore improves computing efficiency. We tested this algorithm on several datasets at three different genomic scales: (1) gene-wide (NOS3, CRP, IL6 PPARA, and TNF), (2) region-wide (a Whitehead Institute inflammatory bowel disease dataset and a UK Graves' disease dataset), and (3) chromosome-wide (chromosome 22) levels. SMET offers geneticists with greater flexibilities in SNP tagging than lossless methods with adjustable haplotype diversity coverage (,). In simulation studies, we found that (1) an initial sample size of 50 individuals (100 chromosomes) or more is needed for htSNP selection; (2) the SNP tagging strategy is considerably more efficient when the underlying block structure is taken into account; and (3) htSNP sets at 80,90% , are more cost-effective than the lossless sets in term of relative power, relative risk ratio estimation, and genotyping efforts. Our study suggests that the novel SMET algorithm is a valuable tool for association tests. Genet. Epidemiol. 29:336,352, 2005. © 2005 Wiley-Liss, Inc. [source] An augmented spatial digital tree algorithm for contact detection in computational mechanicsINTERNATIONAL JOURNAL FOR NUMERICAL METHODS IN ENGINEERING, Issue 2 2002Y. T. Feng Abstract Based on the understanding of existing spatial digital tree-based contact detection approaches, and the alternating digital tree (ADT) algorithm in particular, a more efficient algorithm, termed the augmented spatial digital tree (ASDT) algorithm, is proposed in the present work. The ASDT algorithm adopts a different point representation scheme that uses only the lower corner vertex to represent a (hyper-)rectangle, with the upper corner vertex serving as the augmented information. Consequently, the ASDT algorithm can keep the working space the same as the original n -dimensional space and, in general, a much better balanced tree can be expected. This, together with the introduction of an additional bounding subregion for the rectangles associated with each tree node, makes it possible to significantly reduce the number of node visits in the region search, although each node visit may be slightly more expensive. Three examples arising in computational mechanics are presented to provide an assessment of the performance of the ASDT. The numerical results indicate that the ASDT is, at least, over 3.9 times faster than the ADT. Copyright © 2002 John Wiley & Sons, Ltd. [source] A new approach to automated first-order multiplet analysisMAGNETIC RESONANCE IN CHEMISTRY, Issue 5 2002Sergey Golotvin Abstract The dependence of the values of NMR spin,spin coupling constants on molecular conformation can be a valuable tool in the structure determination process. The continuing increase in the resonance frequency of modern NMR spectrometers allows an increasing number of resonances to be examined using first-order multiplet analysis. While this can easily be done for the simplest patterns (doublets, triplets, quartets), more complex patterns can be extremely difficult to analyze. The task of deducing the coupling constant values from a multiplet is the reverse process of generating a conventional splitting tree from a single line (chemical shift) by sequential branching using a given set of coupling constants. We present a simple, straightforward method of deducing coupling constant values from first-order multiplets based on a general inverted splitting tree algorithm but also including a peak intensity normalization procedure that utilizes multiplet symmetry and generates a set of possible first-order intensity distribution patterns. When combined with an inverted splitting tree algorithm, it is possible to find an intensity pattern that allows the deduction of a proper set of coupling constants. Copyright © 2002 John Wiley & Sons, Ltd. [source] Toward faster algorithms for dynamic traffic assignment.NETWORKS: AN INTERNATIONAL JOURNAL, Issue 1 2003Abstract Being first in a three-part series promising a practical solution to the user-equilibrium dynamic traffic assignment problem, this paper devises a parametric quickest-path tree algorithm, whose model makes three practical assumptions: (i) the traversal time of an arc i , j is a piecewise linear function of the arrival time at its i -node; (ii) the traversal time of a path is the sum of its arcs' traversal times; and (iii) the FIFO constraint holds, that is, later departure implies later arrival. The algorithm finds a quickest path, and its associated earliest arrival time, to every node for every desired departure time from the origin. Its parametric approach transforms a min-path tree for one departure-time interval into another for the next adjacent interval, whose shared boundary the algorithm determines on the fly. By building relatively few trees, it provides the topology explicitly and the arrival times implicitly of all min-path trees. Tests show the algorithm running upward of 10 times faster than the conventional brute-force approach, which explicitly builds a min-path tree for every departure time. Besides dynamic traffic assignment, other applications for which these findings have utility include traffic control planning, vehicle routing and scheduling, real-time highway route guidance, etc. © 2002 Wiley Periodicals, Inc. [source] Modeling above-ground litterfall in eastern Mediterranean conifer forests using fractional tree cover, and remotely sensed and ground dataAPPLIED VEGETATION SCIENCE, Issue 4 2010Sibel Taskinsu-Meydan Abstract Question: How can we model above-ground litterfall in Mediterranean conifer forests using remotely sensed and ground data, and geographic information systems (GIS)? Location: Eastern Mediterranean conifer forest of Turkey. Methods: Above-ground litterfall from Mediterranean forest stands of Pinus nigra, Cedrus libani, Pinus brutia and Juniperus excelsa and mixed Abies cilicica, C. libani and P. nigra was modeled as a function of fractional tree cover using a regression tree algorithm, based on IKONOS and Landsat TM/ETM+data. Landsat TM/ETM+images for the study area were used to map actual stand patterns, based on a land-cover map of species stands using a supervised classification. Results: Total amount of annual above-ground litterfall for the entire study area (12 260 km2) was estimated at 417.2 Mg ha,1 for P. brutia, 291.1 Mg ha,1 for the mixed stand, 115.5 Mg ha,1 for P. nigra, 54.6 Mg ha,1 for J. excelsa and 45.9 Mg ha,1 for C. libani. The maps generated indicate the distribution of the seasonal amount of total above-ground litterfall for different species and the distribution of species stands in the study area. There was an increase in the amount of above-ground litterfall for P. brutia stand in summer, for J. excelsa in autumn and for C. libani, P. nigra and the mixed stand of A. cilicica, P. nigra and C. libani in winter. Conclusion: Application of this model helps to improve the accuracy of estimated litterfall input to soil organic carbon pools in the Mediterranean conifer forests. [source] Prevalence of rheumatoid arthritis in persons 60 years of age and older in the United States: Effect of different methods of case classificationARTHRITIS & RHEUMATISM, Issue 4 2003Elizabeth K. Rasch Objective To determine prevalence estimates for rheumatoid arthritis (RA) in noninstitutionalized older adults in the US. Prevalence estimates were compared using 3 different classification methods based on current classification criteria for RA. Methods Data from the Third National Health and Nutrition Examination Survey (NHANES-III) were used to generate prevalence estimates by 3 classification methods in persons 60 years of age and older (n = 5,302). Method 1 applied the "n of k" rule, such that subjects who met 3 of 6 of the American College of Rheumatology (ACR) 1987 criteria were classified as having RA (data from hand radiographs were not available). In method 2, the ACR classification tree algorithm was applied. For method 3, medication data were used to augment case identification via method 2. Population prevalence estimates and 95% confidence intervals (95% CIs) were determined using the 3 methods on data stratified by sex, race/ethnicity, age, and education. Results Overall prevalence estimates using the 3 classification methods were 2.03% (95% CI 1.30,2.76), 2.15% (95% CI 1.43,2.87), and 2.34% (95% CI 1.66,3.02), respectively. The prevalence of RA was generally greater in the following groups: women, Mexican Americans, respondents with less education, and respondents who were 70 years of age and older. Conclusion The prevalence of RA in persons 60 years of age and older is ,2%, representing the proportion of the US elderly population who will most likely require medical intervention because of disease activity. Different classification methods yielded similar prevalence estimates, although detection of RA was enhanced by incorporation of data on use of prescription medications, an important consideration in large population surveys. [source] A perception-driven autonomous urban vehicleJOURNAL OF FIELD ROBOTICS (FORMERLY JOURNAL OF ROBOTIC SYSTEMS), Issue 10 2008John Leonard This paper describes the architecture and implementation of an autonomous passenger vehicle designed to navigate using locally perceived information in preference to potentially inaccurate or incomplete map data. The vehicle architecture was designed to handle the original DARPA Urban Challenge requirements of perceiving and navigating a road network with segments defined by sparse waypoints. The vehicle implementation includes many heterogeneous sensors with significant communications and computation bandwidth to capture and process high-resolution, high-rate sensor data. The output of the comprehensive environmental sensing subsystem is fed into a kinodynamic motion planning algorithm to generate all vehicle motion. The requirements of driving in lanes, three-point turns, parking, and maneuvering through obstacle fields are all generated with a unified planner. A key aspect of the planner is its use of closed-loop simulation in a rapidly exploring randomized trees algorithm, which can randomly explore the space while efficiently generating smooth trajectories in a dynamic and uncertain environment. The overall system was realized through the creation of a powerful new suite of software tools for message passing, logging, and visualization. These innovations provide a strong platform for future research in autonomous driving in global positioning system,denied and highly dynamic environments with poor a priori information. © 2008 Wiley Periodicals, Inc. [source] | ||||||||||