UID:
edoccha_9961421174702883
Format:
1 online resource (378 pages)
Edition:
First edition.
ISBN:
0-443-15929-7
,
9780443159299
Series Statement:
Woodhead Publishing Series in Civil and Structural Engineering Series
Note:
Intro -- Data Analysis in Pavement Engineering -- Copyright -- Contents -- Preface -- Introduction -- Chapter 1: Pavement performance data -- Chapter outline -- 1.1. Introduction -- 1.2. Pavement performance indices -- 1.2.1. Development of pavement performance indices -- 1.2.2. Pavement performance indices in China -- 1.3. Pavement management system -- 1.4. Pavement performance models -- 1.4.1. Classic pavement performance models -- 1.4.2. Time-performance models -- 1.5. The LTPP database -- 1.5.1. The LTPP program -- 1.5.2. Asphalt pavement performance data in LTPP -- 1.6. Data analysis in pavement engineering -- 1.6.1. An overview -- 1.6.2. Machine learning methods -- 1.6.3. Summary -- Questions -- References -- Chapter 2: Fundamentals of statistics -- Chapter outline -- 2.1. Introduction -- 2.2. Random variables -- 2.2.1. Discrete random variables -- 2.2.2. Continuous random variables -- 2.2.3. Joint distribution -- 2.3. Statistical descriptions of data -- 2.3.1. Sampling methods -- 2.3.2. Numerical summaries -- 2.3.3. Graphical summaries -- 2.3.4. Covariance and correlation -- 2.4. Functions of normal distributions -- 2.4.1. Distributions of the sample mean and variance -- 2.4.2. Compare two sample means and variance -- 2.5. Statistical inference -- 2.5.1. Point estimate -- 2.5.2. Interval estimate -- 2.6. Hypothesis tests -- 2.6.1. Concepts and procedures -- 2.6.2. One-tailed and two-tailed tests -- 2.6.3. Tests for one sample -- 2.6.4. Tests for two samples -- 2.6.5. Proportion tests -- 2.7. Case 1: Significance test of concrete strength -- 2.7.1. Background and data -- 2.7.2. Discussion of results -- Questions -- References -- Chapter 3: Design of experiments -- Chapter outline -- 3.1. Introduction -- 3.1.1. Design of experiments -- 3.1.2. Analysis of variance -- 3.2. Design of experiments -- 3.2.1. Definition -- 3.2.2. Principles.
,
3.2.3. Types of experimental design -- 3.3. Analysis of variance -- 3.3.1. Definitions and procedures -- 3.3.2. Assumptions -- 3.4. One-way ANOVA -- 3.4.1. Procedures -- 3.4.2. Multiple comparisons -- 3.5. Two-way ANOVA -- 3.5.1. Two-way ANOVA without interaction -- 3.5.2. Two-way ANOVA with interaction -- 3.6. Orthogonal design -- 3.6.1. Orthogonal design without interaction -- 3.6.2. Orthogonal design with interaction -- 3.7. Case 1: Concrete strength test analysis through the two-way ANOVA with interaction -- 3.7.1. Background and data -- 3.7.2. Discussion of results -- 3.8. Case 2: Pavement treatment evaluation through orthogonal design with interaction -- 3.8.1. Background and data -- 3.8.2. Discussion of results -- Questions -- References -- Chapter 4: Regression -- Chapter outline -- 4.1. Introduction -- 4.1.1. Multiple linear regression -- 4.1.2. Nonlinear regression -- 4.1.3. Clusterwise regression -- 4.1.4. MARS -- 4.1.5. LASSO -- 4.1.6. Fuzzy logic -- 4.2. SLR -- 4.2.1. Model Definition -- 4.2.2. Parameter estimate -- 4.2.3. Significance test -- 4.3. MLR -- 4.3.1. Model definition -- 4.3.2. Parameter estimate -- 4.3.3. Significance test -- 4.4. Linear regression diagnostics -- 4.4.1. Model assumptions -- 4.4.2. Residual diagnostics -- 4.4.3. Multicollinearity -- 4.4.4. Box-Cox transformation -- 4.5. Stepwise regression -- 4.5.1. Procedures -- 4.5.2. Variable selection -- 4.6. Polynomial regression -- 4.7. Nonlinear regression -- 4.8. Case study 1: Pavement maintenance effectiveness evaluation -- 4.8.1. Background and data -- 4.8.2. Discussion of results -- 4.9. Case 2: Pavement roughness prediction using linear regression with interaction -- 4.9.1. Background and data -- 4.9.2. Discussion of results -- 4.10. Case 3: Pavement roughness prediction using nonlinear regression -- Questions -- References -- Chapter 5: Logistic regression.
,
Chapter outline -- 5.1. Introduction -- 5.1.1. Binary logistic regression -- 5.1.2. Ordinal and multinomial logistic regression -- 5.1.3. Ordinal probit model -- 5.2. Binary logistic regression -- 5.2.1. Model -- 5.2.2. Parameter estimate -- 5.2.3. Model fitness -- 5.3. Multinomial logistic regression -- 5.4. Ordinal logistic regression -- 5.5. Generalized linear model -- 5.6. Case: Cracking probability of pavement maintenance treatments -- 5.6.1. Background and data -- 5.6.2. Discussion of results -- Questions -- References -- Chapter 6: Count data models -- Chapter outline -- 6.1. Introduction -- 6.1.1. Poisson regression model -- 6.1.2. Zero-inflated model -- 6.2. Poisson regression -- 6.2.1. Model -- 6.2.2. Parameter estimate -- 6.2.3. Model fitness -- 6.3. Negative binomial model -- 6.4. Zero-inflated model -- 6.4.1. Zero-inflated Poisson regression -- 6.4.2. Zero-inflated negative binomial model -- 6.4.3. Parameter estimate and model fitness -- 6.5. Case: Development of pavement transverse cracks -- 6.5.1. Background and data -- 6.5.2. Discussion of results -- Questions -- References -- Chapter 7: Survival analysis -- Chapter outline -- 7.1. Introduction -- 7.1.1. Nonparametric models -- 7.1.2. Semiparametric models -- 7.1.3. Parametric models -- 7.2. Data censoring -- 7.3. Survival functions -- 7.4. Nonparametric model -- 7.4.1. Kaplan-Meier method -- 7.4.2. Life table -- 7.4.3. Log-rank test -- 7.5. Semiparametric models -- 7.6. Parametric models -- 7.6.1. Exponential survival model -- 7.6.2. Weibull survival model -- 7.7. Case: The parametric survival model of pavement patching -- 7.7.1. Background and data -- 7.7.2. Discussion of results -- Questions -- References -- Chapter 8: Time series -- Chapter outline -- 8.1. Introduction -- 8.2. Definition and decomposition -- 8.3. Moving average smoothing -- 8.3.1. Simple moving average.
,
8.3.2. Weighted moving average -- 8.3.3. Quadratic moving average -- 8.4. Exponential smoothing -- 8.4.1. Single exponential smoothing -- 8.4.2. Double exponential smoothing -- 8.4.3. Triple exponential smoothing -- 8.5. Parametric models -- 8.5.1. AR model -- 8.5.2. MA model -- 8.5.3. ARMA model -- 8.5.4. ARIMA model -- 8.5.5. Lag operator -- 8.6. Multivariate time series models -- 8.6.1. Vector time series model -- 8.6.2. ARIMA with exogenous inputs -- 8.7. Case: Time series models for pavement roughness prediction -- 8.7.1. ARIMA model -- 8.7.2. VAR model -- 8.7.3. ARIMAX model -- Questions -- References -- Chapter 9: Stochastic process -- Chapter outline -- 9.1. Introduction -- 9.1.1. Markov chain model -- 9.1.2. Gaussian process regression -- 9.2. Stochastic process -- 9.3. Markov chain -- 9.4. Homogeneity -- 9.5. Stationary distribution -- 9.6. Finite state Markov chain -- 9.7. Transition probability matrix based on an ordered probit model -- 9.8. Case 1: Pavement performance prediction using a homogeneous Markov chain -- 9.9. Case 2: Pavement performance prediction based on a dynamic Markov chain -- 9.9.1. Background and data -- 9.9.2. Discussion of results -- Questions -- References -- Chapter 10: Decision trees and ensemble learning -- Chapter outline -- 10.1. Introduction -- 10.1.1. Decision trees -- 10.1.2. Ensemble learning -- 10.2. Decision tree models -- 10.2.1. Structure -- 10.2.2. Hunts algorithm -- 10.2.3. Overfitting -- 10.3. ID3 -- 10.3.1. Entropy -- 10.3.2. Algorithm -- 10.4. C4.5 -- 10.5. CART -- 10.5.1. Discrete response -- 10.5.2. Continuous response -- 10.6. Ensemble learning -- 10.7. Case: Classifying pavement patching serviceability -- 10.7.1. Background and data -- 10.7.2. Discussion of results -- Questions -- References -- Chapter 11: Neural networks -- Chapter outline -- 11.1. Introduction -- 11.1.1. NN -- 11.1.2. DNN.
,
11.1.3. CNN -- 11.1.4. RNN -- 11.2. Single-layer neural network -- 11.2.1. McCulloch-Pitts neuron -- 11.2.2. Model structure -- 11.3. Two-layer neural network -- 11.3.1. Model structure -- 11.3.2. Activation function -- 11.3.3. Training of NNs -- 11.4. Multilayer neural network -- 11.5. Deep learning -- 11.6. Convolutional neural network -- 11.6.1. Convolution layer -- 11.6.2. Activation layer -- 11.6.3. Pooling layer -- 11.6.4. Fully connected layer -- 11.6.5. Training of A CNN -- 11.7. Computation of neural network -- 11.8. Case 1: Pavement roughness prediction -- 11.9. Case 2: Climatic region classification -- Questions -- References -- Chapter 12: Support vector machine and k-nearest neighbors -- Chapter outline -- 12.1. Introduction -- 12.2. Algorithm -- 12.2.1. Definition -- 12.2.2. Hyperplane -- 12.2.3. Constraints -- 12.2.4. Lagrangian dual form -- 12.2.5. Kernel function -- 12.2.6. Advantages of SVM -- 12.3. k-nearest neighbors -- 12.4. Performance metrics of classification models -- 12.5. Case study: Pavement roughness prediction based on SVM and k-NN -- 12.5.1. Background and data -- 12.5.2. Discussion of results -- Questions -- References -- Chapter 13: Principal component analysis -- Chapter outline -- 13.1. Introduction -- 13.2. Definition -- 13.3. Method -- 13.3.1. Model -- 13.3.2. Computation of coefficients -- 13.4. Procedure -- 13.5. Case: PCA of climatic data -- 13.5.1. Background and data -- 13.5.2. Discussion of results -- Questions -- References -- Chapter 14: Factor analysis -- Chapter outline -- 14.1. Introduction -- 14.2. Model -- 14.2.1. Definition -- 14.2.2. Assumptions -- 14.2.3. Contributions of factors -- 14.3. Calculation of factors -- 14.3.1. Principal component method -- 14.3.2. Principal factor method -- 14.3.3. Maximum likelihood estimation method -- 14.4. Factor rotation -- 14.5. Factor score.
,
14.6. Case 1: Climatic data analysis.
Additional Edition:
ISBN 9780443159282
Language:
English
Bookmarklink