Results 1 - 10 of 21,174

Table 3 Results for the detection of density-independent Malthusian demography. The val- ues in each cell represent the percentage a given model is chosen by means of SRM.

in VC-dimension and Structural Risk Minimization for the analysis of nonlinear ecological models
by Giorgio Corani, Marino Gatto
"... In PAGE 14: ... In order to be coherent with the VC estimates experimental settings, identification algorithms of nonlinear models are initialized 40 times. Table3 shows a summary of the results; it does not report the details regarding all the 24 different simulation settings, rather it provides the average model choice percentages for wide classes of settings. In the global average, the SRM test correctly recognizes the Malthusian model about 78% of the times; the remaining 22% of the times it chooses Beverton- Holt, which is the nonlinear model with the lowest VC-dimension.... ..."

Table 2: Detailed results for the recognition of the simple Malthusian model with different settings. Percentages in bold refer to the model that really generated the data (the higher, the better). See text for symbols.

in Model Selection in Demographic Time Series Using
by Vc-Bounds Giorgio Corani, Giorgio Corani, Marino Gatto
"... In PAGE 12: ... The average is taken over all the different simulation settings (model parameters, simulation length, noise level). More detailed results are reported in Table2 , 3, 4, 5, 6, that evidence the sensitivity of the recognition success to variations of each single parameter, by pooling the experiments that share the same value of a given parameter. For instance, the row (a = 0:5) in Table 2 refers to the average result obtained on 8000 = 500 4 (different values of n) 4 (different values of q) experiments.... In PAGE 12: ... More detailed results are reported in Table 2, 3, 4, 5, 6, that evidence the sensitivity of the recognition success to variations of each single parameter, by pooling the experiments that share the same value of a given parameter. For instance, the row (a = 0:5) in Table2 refers to the average result obtained on 8000 = 500 4 (different values of n) 4 (different values of q) experiments. As for the Malthusian model, it is almost always recognized (see Table 1) both by SIC and SRM (99% and 98%), while FPE fails about 20% of times, selecting a density-dependent demography.... In PAGE 12: ... As for the Malthusian model, it is almost always recognized (see Table 1) both by SIC and SRM (99% and 98%), while FPE fails about 20% of times, selecting a density-dependent demography. Looking at the detailed results ( Table2 ), one notes that SIC and SRM recognize correctly the Malthusian demography, as they are in practice insensitive to any variation of the noise level n, the simulation length q or the drift parameter a; however, FPE too shows little sensitivity of its performances to changes in one of these parameters. As for the ability to recognize the Ricker model (simulations started at N0 = 100), SRM (92%) strongly outperforms FPE and SIC (78% and 72% respectively); for whatever value of a, n and q, a consistent advantage of SRM over both FPE and SIC is found (see Table 3).... ..."

Table 1: Comparison of tracking results

in On-Line Density-Based Appearance Modeling for Object Tracking
by Bohyung Han Larry

Table 1. Nuclear Features Measured from Cervical Cells

in Detection Of Malignancy Associated Changes In Thionin Stained Cervical Cells
by Jennifer Hallinan And, Cervical Cells, Jennifer Hallinan, Paul Jackway
"... In PAGE 3: ....2. Features A total of forty-two features were extracted from each nucleus. These are features which have been described in the literature as being more or less useful for MAC analysis ( Table1 .).... ..."

Table 2: Features that influence hospital mortality according to CART and stepwise logistic regression analysis

in Prediction of Outcome in the Critically Ill Using an Artificial Neural Network Synthesised By a Genetic Algorithm
by Richard Dybowski, Peter Weller, René Chang, Ren� Chang, Vanya Gant 1996
Cited by 2

Table 3.1 Correlation coefficients between line flows and MVAr margin

in for Transfer Limits
by Final Project Report, Anjan Bose, Robert Stuart, Ben Williams, Mark Willis, Liqiang Chen, Mohammad Vaziri

TABLE I BASE FEATURE SET

in Sensor Selection for Maneuver Classification
by Kari Torkkola, Srihari Venkatesan, Huan Liu 2004
Cited by 2

TABLE I BASE FEATURE SET

in Sensor Sequence Modeling for Driving
by Kari Torkkola, Srihari Venkatesan, Huan Liu

Table 4 Saccadic frequency and saccadic bias towards single stimulus features in the single-feature condition and towards pairs of features in the two-feature condition

in Peripheral and Parafoveal Cueing and Masking Effects on Saccadic Selectivity in a Gaze-Contingent Window Paradigm
by Marc Pomplun, Eyal M. Reingold, Jiye Shen 2001

Table 1. Characteristics of building datasets used to derive NBSD (listed alphabetically).

in 2006: Emerging urban databases for meteorological and dispersion
by Steven J. Burian, Michael J. Brown, Timothy N. Mcpherson, James Hartman, Woosuk Han, Indumathi Jeyach, Johnathan F. Rush
"... In PAGE 4: ... Metropolitan areas included in the tall building district database. The characteristics of the building datasets used to derive the NBSD are contained in Table1 . All data extents are smaller than the complete metropolitan area, but are centered on the important tall building districts.... ..."
Cited by 1
Results 1 - 10 of 21,174