1,721,043 research outputs found
shape_variation_ADNI
No full text2015 evaluation of shape variation in ADNI brains, conducted by Elias Chaibub Neto
shape_variation_Mindboggle101
No full text2014 analysis of shape variation in the Mindboggle-101 data with Elias Chaibub Neto. Includes shape measures computed on the manual labels and the mindboggle labels assigned to the Mindboggle-101 brain images
Comparison between the data re-sampling approaches and the vectorized multinomial sampling bootstrap, in the American law school data.
No full text<p>Panel a compares the approach’s time expenditures. Panels b, c and d compare, respectively, the computation time ratio of the “for loop”, “R/bootstrap” and “R/boot” approaches against the vectorized bootstrap. The bottom panels present the </p><p></p><p></p><p></p><p></p><p><mi>θ</mi><mo>^</mo></p><mo>*</mo><p></p><mo>=</mo><p>c</p><p>o<mo>^</mo></p>r<p></p><mo stretchy="false">(</mo><p>LSAT<mo>*</mo></p><mo>,</mo><p>GPA<mo>*</mo></p><mo stretchy="false">)</mo><p></p><p></p><p></p> distributions generated with <i>B</i> = 1,000,000.<p></p
Speeding Up Non-Parametric Bootstrap Computations for Statistics Based on Sample Moments in Small/Moderate Sample Size Applications.
No full textIn this paper we propose a vectorized implementation of the non-parametric bootstrap for statistics based on sample moments. Basically, we adopt the multinomial sampling formulation of the non-parametric bootstrap, and compute bootstrap replications of sample moment statistics by simply weighting the observed data according to multinomial counts instead of evaluating the statistic on a resampled version of the observed data. Using this formulation we can generate a matrix of bootstrap weights and compute the entire vector of bootstrap replications with a few matrix multiplications. Vectorization is particularly important for matrix-oriented programming languages such as R, where matrix/vector calculations tend to be faster than scalar operations implemented in a loop. We illustrate the application of the vectorized implementation in real and simulated data sets, when bootstrapping Pearson's sample correlation coefficient, and compared its performance against two state-of-the-art R implementations of the non-parametric bootstrap, as well as a straightforward one based on a for loop. Our investigations spanned varying sample sizes and number of bootstrap replications. The vectorized bootstrap compared favorably against the state-of-the-art implementations in all cases tested, and was remarkably/considerably faster for small/moderate sample sizes. The same results were observed in the comparison with the straightforward implementation, except for large sample sizes, where the vectorized bootstrap was slightly slower than the straightforward implementation due to increased time expenditures in the generation of weight matrices via multinomial sampling
Comparison of the time spent in the generation of the bootstrap weights matrix, <i><b>W</b></i>*, versus the time spent on all other matrix/vector operations involved in the calculation of the vector of bootstrap replications.
No full text<p>Panel a reports the time (in seconds) spent in the generation of <i><b>W</b></i>* (blue curve) and on the remaining matrix operations (red curve) as a function of sample sizes varying from 15 to 915, when <i>B</i> = 10,000. Panels b and c present analogous results for <i>B</i> equal to 100,000, and 1,000,000, respectively. Panel d presents the ratio of the time spent in the generation of <i><b>W</b></i>* against all other matrix/vector operations, for <i>B</i> equal to 10,000, 100,000, and 1,000,000.</p
Comparison of the bootstrap implementations, when bootstrapping Pearson’s sample correlation, θ^*=co^r(x1*,x2*), for sample sizes varying from 15 to 915.
No full text<p>The data was simulated according to <i>x</i><sub>1<i>i</i></sub> = <i>ϵ</i><sub>1<i>i</i></sub>, <i>x</i><sub>2<i>i</i></sub> = <i>x</i><sub>1<i>i</i></sub>+<i>ϵ</i><sub>2<i>i</i></sub>, <i>ϵ</i><sub><i>ji</i></sub> ∼ N(0,1), <i>j</i> = 1,2. Results based on 10,000 (panels a-d), 100,000 (panels e-h), and 1,000,000 (panels i-l) bootstrap replications. The left panels (a, e, and i) compare the approach’s time expenditures. The remaining panels show the computation time ratios (in log scale) comparing the resampling approaches against the vectorized implementation. The horizontal line is set at zero and represents the threshold below which the data resampling approach outperforms the vectorized implementation.</p
Comparison between the data re-sampling approaches and the vectorized multinomial sampling bootstrap, in a subset of the American law school data.
No full text<p>Panel a compares the approach’s time expenditures. Panels b, c and d compare, respectively, the computation time ratio of the “for loop”, “R/bootstrap” and “R/boot” approaches against the vectorized bootstrap. The bottom panels present the </p><p></p><p></p><p></p><p></p><p><mi>θ</mi><mo>^</mo></p><mo>*</mo><p></p><mo>=</mo><p>c</p><p>o<mo>^</mo></p>r<p></p><mo stretchy="false">(</mo><p>LSAT<mo>*</mo></p><mo>,</mo><p>GPA<mo>*</mo></p><mo stretchy="false">)</mo><p></p><p></p><p></p> distributions generated with <i>B</i> = 1,000,000.<p></p
Going Beyond Counting First Authors in Author Co-citation Analysis
Get PDFThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
- …
