Juan Jiménez-Sánchez (10125734), Álvaro Martínez-Rubio (10125737), Anton Popov (10125740) et al. · Figshare · 2021
<p><b>A)</b> Basic algorithm (for implementation purposes). Initialization requires creating a 3D grid, specifying the final simulation time, defining subpopulation traits and setting the initial state. Temporal iterations are then carried out until the end time is reached. At each time step and each voxel, all clonal populations are updated. This updating involves calculating how many cells will proliferate, migrate, mutate or die. When all populations at all voxels have been evaluated, they are updated synchronously. <b>B)</b> Two-dimensional example of model behavior. Synchronous updating results in population increasing, diversifying and spreading at each time step, with probabilities computed according to the biological rules. Compartment color indicates occupancy. Cell color indicates cell type. <b>C)</b> Bottom image is a slice of an actual simulation, with colors indicating occupation. Each voxel contains a variable number of cells and subtypes as depicted above.</p>
Cited by 0Open access
Steve Hoffmann (242470), Christian Otto (315144), Stefan Kurtz (1227) et al. · Figshare · 2015
Cited by 0Open access
Tom Willaert · VUBIR (Vrije Universiteit Brussel) · 2017
Cited by 0
Laura Martignon · Elsevier eBooks · 2015
Cited by 1
M. Butta, P. Ripka, J. Kubík · 2014
Cited by 0
Mirela Voicu, Gabriela Mircea · 2015
Cited by 0
Joel Susskind, John Blaisdell, Lena Iredell · 2016
AIRS was launched on EOS Aqua on May 4, 2002 together with ASMU-A and HSB to form a next generation polar orbiting infrared and microwave atmosphere sounding system (Pagano et al 2003). The theoretical approach used to analyze AIRS/AMSU/HSB data in the presence of clouds in the AIRS Science Team Version 3 at-launch algorithm, and that used in the Version 4 post-launch algorithm, have been published previously. Significant theoretical and practical improvements have been made in the analysis of AIRS/AMSU data since the Version 4 algorithm. Most of these have already been incorporated in the AIRS Science Team Version 5 algorithm (Susskind et al 2010), now being used operationally at the Goddard DISC. The AIRS Version 5 retrieval algorithm contains three significant improvements over Version 4. Improved physics in Version 5 allowed for use of AIRS clear column radiances in the entire 4.3 m CO2 absorption band in the retrieval of temperature profiles T(p) during both day and night. Tropospheric sounding 15 m CO2 observations were used primarily in the generation of clear column radiances for all channels. This new approach allowed for the generation of accurate Quality Controlled values of and T(p) under more stressing cloud conditions. Secondly, Version 5 contained a new methodology to provide accurate case-by-case error estimates for retrieved geophysical parameters and for channel-by-channel clear column radiances. Thresholds of these error estimates are used in a new approach for Quality Control. Finally, Version 5
Cited by 0
Encyclopedia of Statistical Sciences · 2004
Cited by 0
Saul I. Gass, Michael C. Fu · 2013
Cited by 0
Yadin Dudai · 2002
Abstract ‘Begin at the beginning’, said the King of Hearts, ‘and go on till you come to the end: then stop’. He thus provided White Rabbit with an algorithm (Carroll 1865). The term ‘algorithm’ is derived from Latinization of the name of one of the most creative mathematician in medieval Islam, Al-Kwarizmi (780–c. 850; Boyer 1989; Colish 1997). In modern times algorithmic is a field fundamental to the science of computing (Harel 1987). In the neurosciences algorithms are encountered in multiple contexts (Marr 1982; Hinton 1989; Church land and Sejnowski 1992). One of these is in *models of biological learning.
Cited by 0
Luciani, Annie · HAL (Le Centre pour la Communication Scientifique Directe) · 2007
International audience
Cited by 0Open access
Corrado Coia, Brian J. Ross, Corrado Coia et al. · 2009
Granular synthesis is a form of audio synthesis that consists of breaking audio into tiny millisecond chucks. This paper describes the EGDE system. EGDE (Evolutionary Granular Delay Environment) is a software plug-in that permits its granular delay effects to be time synchronized with the host application. Its interface features an interactive genetic algorithm, to be used for user exploration of granular synthesis parameters. Interactive evolution is ideal in this application, as it permits a user to interactively explore a wide variety of effects that arise with different combinations of the granular delay parameters. A goal of EGDE is to provide a granular synthesis interface with an intuitive and efficient means of auditioning and rating effect parameters, while minimizing user exhaustion. This is particularly important with a granular synthesis, since many parameter settings will be undesirable to most listeners. Typically, each parameter set in the population can be auditioned in a matter of a few seconds or less. A minimalist 3-value evaluation scheme lets the user either clone, use, or delete candidate parameter sets. Mutation and crossover rates can also be fine tuned as desired. The interface lets the user directly tweak parameters, thus permitting direct user editing of chromosomes. The evolutionary interface could be easily adapted to other musical applications in the future, for example, generalized synthesis engines.
Cited by 0
Mathieu Bouillaguet, Luciana Arantes, Pierre Sens et al. · 2009
Cited by 0
Amy Greenwald, Zheng Li, Warren Schudy et al. · 2008
Cited by 0
Mingyan Jiang, Francisco Rubio, Yong Wang et al. · 2008
Vol.:-No.:-Date: 20-21 2007 This publication has been included here just to facilitate downloads to those people asking for personal use copies. This material may be published at copyrighted journals or conference proceedings, so personal use of the download is required. In particular, publications from IEEE have to be downloaded according to the following IEEE note: c○2007 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.
Cited by 0
Jonathan Côté (527819), Yvette Bonvalot (527820), Gaétan Carrier (527821) et al. · Figshare · 2015
<p>Algorithm for the determination of parameters values of the model with experimental data of urine excretion profile.</p>
Cited by 0Open access
Daniel Rivas-Barragan (9721575), Sarah Mubeen (6704834), Francesc Guim Bernat (9721578) et al. · Figshare · 2020
Cited by 0Open access
Dario Martelli (541255), Fiorenzo Artoni (541256), Vito Monaco (541257) et al. · Figshare · 2015
<p>The scheme shows the entire algorithm used in this study. The red, green and blue boxes represent respectively the inner <i>LOOCV</i>, the outer <i>LOOCV</i> steps and the repetitions for all the subjects. The gray boxes represent the main core of the algorithm (i.e., <i>ICA1</i>, <i>ICA2</i> and <i>NN</i>). The two switches are rigidly interconnected (dotted line) and allow the flow of information coming from the upper layers of the algorithm to go through either the purple lines (i.e., it occurs when the algorithm processes the dataset accounting for all body segments; see the section titled <i>Processing of the datasets accounting for all segments</i>) or the blue lines (i.e., it occurs when the algorithm processes the dataset accounting for a subset of all body segments; see the section titled <i>Processing of the datasets accounting for a subset of all segments</i>). Note that when the switches are commuted toward the purple (blue) lines no information can go through the blue (purple) lines, that is, the unselected lines are neglected by the algorithm. <i>M</i> is the initial dataset constructed by concatenating data related to the 20 collected trials and is preliminarily parsed out by <i>ICA1</i> to determine the optimal number <i>N</i> of retained <i>ICs</i> and the <i>Ranking</i> concerning their informativeness. These information were respectively used to run the <i>ICA2</i> either on the dataset accounting for all body segments (purple line) or on that accounting for a limited number of segments (light blue line). In this regard the box <i>F</i> selects the subset of segments being processed according to their <i>Ranking</i>. Then, 19 out of 20 trials (Training Set–<i>TrS<sub>i</sub></i>) are selected to train the model, whereas the excluded trial (Test Set–<i>TeS<sub>i</sub></i>) is used to test its performance. This procedure is repeated 20 times (“outer” <i>LOOCV</i>) leaving a different trial as <i>TeS<sub>i</sub></i> each time (i.e., 20 <i>TrS<sub>i</sub></i> and 20 <i>TeS<sub>i</sub></i> for each subject– see green box). In order to train the model, <i>ICA2</i> is performed on the <i>TrS<sub>i</sub></i> with the number of components previously determined. <i>ICA2</i> produces a time-invariant matrix (<i>W<sub>ICAi</sub></i>) which is used to project both the <i>TeS<sub>i</sub></i> and the <i>TrS<sub>i</sub></i> into the independent component space. The resultant projected <i>TrS<sub>i</sub></i> (<i>S<sub>i</sub></i>) is passed on to the Neural Network (<i>NN</i>). A further <i>LOOCV</i> step (inner <i>LOOCV</i>) is applied to <i>S<sub>i</sub></i> (19 trials): 18 of them (<i><sup>j</sup>S<sub>i</sub></i>) are selected to train the <i>NN</i>, whereas the excluded trial (Validation Set – <i><sup>j</sup>VaS<sub>i</sub></i>) is adopted to determine the optimal number of iterations. For each inner <i>LOOCV</i> iteration, the outputs of the <i>NN</i> box are the Bias (<i><sup>j</sup>b<sub>i</sub></i>) and the Weight matrices of the hidden (<i><sup>j</sup>W<sub>hi</sub></i>) and output (<i><sup>j</sup>W<sub>oi</sub></i>) layers, which are later used on the projected <i>TeS<sub>i</sub></i> (<i>S′<sub>i</sub></i>) to calculate the resultant vector <i><sup>j</sup>out<sub>i</sub></i>. Each time point of <i><sup>j</sup>out<sub>i</sub></i> is a dimensionless either <i>W</i>, <i>P</i>, or <i>N/A</i> (uncertainty) class. A potential perturbation is considered as detected if 5 consecutive time points are classified as <i>P</i>. A detected potential perturbation is then labelled as True Positive (<i><sup>j</sup>TP<sub>i</sub></i>), False Negative (<i><sup>j</sup>FN<sub>i</sub></i>), False Positive or True Negative (<i><sup>j</sup>TN<sub>i</sub></i>). Only in the <i><sup>j</sup>TP<sub>i</sub></i> case, the Detection Time (<i><sup>j</sup>DT<sub>i</sub></i>) is defined as the time elapsed from the delivery of the perturbation to the subject and the actual detection. Finally the Mean Detection Time (<i>MDT</i>) is defined as the average of all <i><sup>j</sup>DT<sub>i</sub></i> and the occurrences of <i><sup>j</sup>TP<sub>i</sub></i>, <i><sup>j</sup>FP<sub>i</sub></i>, <i><sup>j</sup>TN<sub>i</sub></i>, <i><sup>j</sup>FN<sub>i</sub></i> are counted (<i>TP<sub>tot</sub></i>, <i>FP<sub>tot</sub></i>, <i>TN<sub>tot</sub></i>, <i>FN<sub>tot</sub></i>). The <i>Sensitivity</i>, <i>Specificity</i> and <i>Accuracy</i> of the whole process are finally calculated as explained in <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0092037#s2" target="_blank"><i>Materials and Methods</i></a>.</p>
Cited by 0Open access
Sarosh N. Fatakia (195565), Stefano Costanzi (82273), Carson C. Chow (43428) · Figshare · 2015
Cited by 0Open access
Mark C. Neyrinck, Nickolay Y. Gnedin, Andrew J. S. Hamilton · 2004
A crucial step in comparing N-body simulations to the observed galaxy distribution is to identify the possible sites of galaxy formation, called dark-matter haloes, in the simulations. Unfortunately, the concept of a dark-matter halo
Cited by 0