02716cam a22003255a 4500991002325489707536131025s2012 nyu 000 0 eng d9783642324772 (hard cover : alk. paper)b14156374-39ule_instBibl. Dip.le Aggr. Matematica e Fisica - Sez. Fisicaeng530.1523LC QC21.3510:53Sirca, Simon479103Computational methods for physicists :compendium for students /Simon Sirca, Martin HorvatNew York :Springer,2012xx, 715 p :ill. ;24 cmGraduate texts in physicsIncludes bibliographical references and indexBasics of numerical analysis -- Solution of nonlinear equations -- Matrix methods -- Transformations of functions and signals -- Statistical description and modeling of data -- Modeling and analysis of time series -- Initial-value problems for ordinary differential equations -- Boundary-value problems for ordinary differential equations -- Difference methods for one-dimensional partial differential equations -- Difference methods for partial differential equations in more than one dim -- Spectral methods for partial differential equationsThis book helps advanced undergraduate, graduate and postdoctoral students in their daily work by offering them a compendium of numerical methods. The choice of methods pays significant attention to error estimates, stability and convergence issues as well as to the ways to optimize program execution speeds. Many examples are given throughout the chapters, and each chapter is followed by at least a handful of more comprehensive problems which may be dealt with, for example, on a weekly basis in a one- or two-semester course. In these end-of-chapter problems the physics background is pronounced, and the main text preceding them is intended as an introduction or as a later reference. Less stress is given to the explanation of individual algorithms. It is tried to induce in the reader an own independent thinking and a certain amount of scepticism and scrutiny instead of blindly following readily available commercial toolsMathematical physicsPhysicsData processingHorvat, Martinauthorhttp://id.loc.gov/vocabulary/relators/aut732550.b1415637410-11-1425-10-13991002325489707536LE006 510:53 SIR12006000171441le006pE72.75-l- 00000.i1564223910-11-14Computational methods for physicists1443300UNISALENTOle00625-10-13ma -engnyu0005055oam 2200493 450 991078000070332120170523091546.00-12-394614-X(OCoLC)853506317(MiFhGG)GVRL8DED(EXLCZ)99255000000110586320140429d2013 uy 0engurun|---uuuuatxtccrInformatics for materials science and engineering data-driven discovery for accelerated experimentation and application /edited by Krishna Rajan1st ed.Oxford Butterworth-Heinemann2013Oxford :Butterworth-Heinemann,2013.1 online resource (xv, 525 pages) illustrations (some color)Gale eBooksDescription based upon print version of record.0-12-394399-X 1-299-73366-2 Includes bibliographical references and index.Front Cover; Informatics for Materials Science and Engineering; Copyright Page; Contents; Preface: A Reading Guide; Acknowledgment; 1. Materials Informatics: An Introduction; 1. The What and Why of Informatics; 2. Learning from Systems Biology: An "Omics" Approach to Mater; 3. Where Do We Get the Information?; 4. Data Mining: Data-Driven Materials Research; References; 2. Data Mining in Materials Science and Engineering; 1. Introduction; 2. Analysis Needs of Science Applications; 3. The Scientific Data-Mining Process; 4. Image Analysis; 5. Dimension Reduction; 5.1 Feature Selection TechniquesDistance Filter Chi-Squared Filter; Stump Filter; Relief; 5.2 Feature Transformation Techniques; Principal Component Analysis (PCA); Isomap; Locally Linear Embedding (LLE); Laplacian Eigenmaps; 5.3 Comparison of Dimension Reduction Methods; 6. Building Predictive and Descriptive Models; 6.1 Classification and Regression; 6.2 Clustering; 7. Further Reading; Acknowledgments; References; 3. Novel Approaches to Statistical Learning in Materials Science; 1. Introduction; 2. The Supervised Binary Classification Learning Problem; 3. Incorporating Side Information; 4. Conformal Prediction5. Optimal Learning 6. Optimal Uncertainty Quantification; 7. Clustering Including Statistical Physics Approaches; 8. Materials Science Example: The Search for New Piezoelectrics; 9. Conclusion; 10. Further Reading; Acknowledgments; References; 4. Cluster Analysis: Finding Groups in Data; 1. Introduction; 2. Unsupervised Learning; 2.1 Principal Components Analysis; 2.2 Clustering; 3. Different Clustering Algorithms and their Implementations in R; 3.1 Agglomerative Hierarchical; 3.2 K-Means; 3.3 Divisive Hierarchical; 3.4 Partitioning Around Medoids (PAM); 3.5 Fuzzy Analysis (FANNY)4. Validations of Clustering Results 4.1 Dunn Index; 4.2 Silhouette Width; 4.3 Connectivity; 5. Rank Aggregation of Clustering Results; 6. Further Reading; Acknowledgments; References; 5. Evolutionary Data-Driven Modeling; 1. Preamble; 2. The Concept of Pareto Tradeoff; 3. Evolutionary Neural Net and Pareto Tradeoff; 4. Selecting the Appropriate Model in EvoNN; 5. Conventional Genetic Programming; 6. Bi-Objective Genetic Programming; 6.1 BioGP Code; 7. Analyzing the Variable Response in EvoNN and BioGP; 8. An Application in the Materials Area; 9. Further Reading; References6. Data Dimensionality Reduction in Materials Science 1. Introduction; 2. Dimensionality Reduction: Basic Ideas and Taxonomy; 3. Dimensionality Reduction Methods: Algorithms, Advantages, and Disadvantages; 3.1 Principal Component Analysis (PCA); PCA Algorithm; 3.2 Isomap; Isomap Algorithm; 3.3 Locally Linear Embedding; LLE Algorithm; 3.4 Hessian LLE; hLLE Algorithm; 4. Dimensionality Estimators; 5. Software; 5.1 Core Functionality; 5.2 User Interface; 6. Analyzing Two Material Science Data Sets: Apatites and Organic Solar Cells; 6.1 Apatite Data; Dimensionality Estimation6.2 Unraveling Process-Morphology Pathways of Organic Solar Cells using SETDiRMaterials informatics: a 'hot topic' area in materials science, aims to combine traditionally bio-led informatics with computational methodologies, supporting more efficient research by identifying strategies for time- and cost-effective analysis. The discovery and maturation of new materials has been outpaced by the thicket of data created by new combinatorial and high throughput analytical techniques. The elaboration of this ""quantitative avalanche""-and the resulting complex, multi-factor analyses required to understand it-means that interest, investment, and research are revisiMaterials scienceData processingMaterials scienceData processing.620.110285Rajan KrishnaMiFhGGMiFhGGBOOK9910780000703321Informatics for materials science and engineering3704044UNINA