![]()
内容推荐 材料信息学是一门新兴的交叉学科,为在材料基因组理念下加速材料科学研究和技术发展提供了一个全新的方法。作为材料和力学学者,作者在推动材料信息学发展方面做了大量工作,在人工智能(AI)、机器学习(ML)和材料科学技术融合交叉方面,有诸多的尝试和心得体会。作者旨在写一本易懂的材料信息学简介,以进一步推动材料信息学的发展。为便于读者尽快理解和掌握材料信息学的核心内容,兼顾成书的完整性,本书分为上下两卷,上卷侧重于机器学习基础,下卷侧重于深度学习并综述材料信息学的现状及发展前景。 本上卷共十二章,内容包括线性回归与线性分类、支持向量机、决策树和K近邻(KNN)、集成学习、贝叶斯定理和期望最大化(EM)算法、符号回归、神经网络、隐型马尔可夫链、数据预处理和特征选择、可解释性机器学习,等等。叙述力求从简单明了的数学定义和物理图像出发,密切结合材料科学研究案例,给出了各种算法的详细步骤,便于读者学习和运用。 本卷为机器学习基础的入门工具性读物,适合理工科科研人员和工程技术工作者快速且轻松地学习、理解和掌握材料信息学的AI和ML方法,也特别适合作为研究生和高年级本科生的教材。 目录 Foreword Preface Symbols and Notations Chapter 1 Introduction References Chapter 2 Linear Regression 2.1 Least Squares Linear Regression 2.2 Principal Component Analysis and Principal Component Regression 2.3 Least Absolute Shrinkage and Selection Operator (L1) 2.4 Ridge Regression (L2) 2.5 Elastic Net Regression 2.6 Multiply Task LASSO (MultiTaskLASSO) Homework References Chapter 3 Linear Classification 3.1 Perceptron 3.2 Logistic Regression 3.3 Linear Discriminant Analysis Homework References Chapter 4 Support Vector Machine 4.1 SVC 4.2 Kernel Functions 4.3 Soft Margin 4.4 SVR Homework References Chapter 5 Decision Tree and K-Nearest-Neighbors (KNN) 5.1 Classification Trees 5.2 Regression Tree 5.3 K-Nearest-Neighbors (KNN) Methods Homework References Chapter 6 Ensemble Learning 6.1 Boosting 6.1.1 AdaBoost 6.1.2 Gradient Boosting Machine (GBM) 6.1.3 eXtreme Gradient Boosting (XGBoost) 6.2 Bagging Homework References Chapter 7 Bayesian Theorem and Expectation-Maximization (EM) Algorithm 7.1 Bayesian Theorem 7.2 Naive Bayes Classifier 7.3 Maximum Likelihood Estimation 7.3.1 Gaussian distribution 7.3.2 Weibull distribution 7.4 Bayesian Linear Regression 7.5 Expectation-Maximization (EM) Algorithm 7.5.1 Gaussian mixture model (GMM) 7.5.2 The mixture of Lorentz and Gaussian distributions 7.6 Gaussian Process (GP) Regression Homework References Chapter 8 Symbolic Regression 8.1 Overview of Evolutionary Computation 8.2 Genetic Programming 8.3 Grammar-Guided Genetic Programming and Grammatical Evolution 8.4 The Application of LASSO in Symbolic Regression Homework References Chapter 9 Neural Networks 9.1 Neural Networks and Perceptron 9.2 Back Propagation Algorithm 9.3 Regularization in NNs 9.3.1 L1 regularization 9.3.2 L2 regularization 9.4 Classification NNs 9.4.1 Binary classification 9.4.2 Multiclassification of multiply grades in a category 9.5 Autoencoders 9.5.1 Introduction 9.5.2 Denoising autoencoder 9.5.3 Sparse autoencoder 9.5.4 Variational autoencoder Homework References Chapter 10 Hidden Markov Chains 10.1 Markov Chain 10.2 Stationary Markov Chain 10.3 Markov Chain Monte Carlo Methods 10.3.1 Metropolis Hastings (M-H) algorithm 10.3.2 Gibbs sampling algorithm 10.4 Calculation Methods for the Probability of Observation Sequence 10.4.1 Direct method 10.4.2 Forward method 10.4.3 Backward method 10.5 Estimation of Optimal State Sequence 10.5.1 Direct method 10.5.2 Viterbi algorithm 10.6 Estimation of Intrinsic Parameters—The Baum-Welch Algorithm Homework References Chapter 11 Data Preprocessing and Feature Selection 11.1 Reliable Data, Normals and Anomalies 11.1.1 Local outlier factor 11.1.2 Isolated forest 11.1.3 One-class support vector machine 11.1.4 Support vector data description 11.2 Feature Selection 11.2.1 Filter approach 11.2.2 Wrapper approach 11.2.3 Embedded approach Homework References Chapter 12 Interpretative SHAP Value and Partial Dependence Plot 12.1 SHapley Additive exPlanation value 12.2 The joint SHAP value of two features 12.3 Partial Dependence Plot Homework References Appendix 1 Vector and Matrix A1.1 Definition A1.1.1 Vector A1.1.2 Matrix A1.2 Matrix Algebra A1.2.1 Inverse and transpose A1.2.2 Trace A1.2.3 Determinant A1.2.4 Eigenvalues and eigenvectors A1.2.5 Singular value decomposition (SVD) A1.2.6 Pseudo inverse A1.2.7 Some useful identities A1.3 Matrix Analysis A1.3.1 Derivative of matrix A1.3.2 Derivative of the determinant of a matrix A1.3.3 Derivative of an inverse matrix A1.3.4 Ja |