Machine learning methods in the environmental sciences 1st edition by William W Hsieh ISBN 0521791928‎ 978-0521791922

Machine learning methods in the environmental sciences 1st edition  by William W Hsieh – Ebook PDF Instant Download/Delivery: 0521791928‎, 978-0521791922
Full download Machine learning methods in the environmental sciences 1st edition after payment

Product details:

ISBN 10: 0521791928
ISBN 13:‎ 978-0521791922
Author: William W Hsieh

Machine learning methods originated from artificial intelligence and are now used in various fields in environmental sciences today. This is the first single-authored textbook providing a unified treatment of machine learning methods and their applications in the environmental sciences. Due to their powerful nonlinear modeling capability, machine learning methods today are used in satellite data processing, general circulation models(GCM), weather and climate prediction, air quality forecasting, analysis and modeling of environmental data, oceanographic and hydrological forecasting, ecological modeling, and monitoring of snow, ice and forests. The book includes end-of-chapter review questions and an appendix listing web sites for downloading computer code and data sources. A resources website containing datasets for exercises, and password-protected solutions are available. The book is suitable for first-year graduate students and advanced undergraduates. It is also valuable for researchers and practitioners in environmental sciences interested in applying these new methods to their own work.
Preface Excerpt
Machine learning is a major subfield in computational intelligence (also called artificial intelligence). Its main objective is to use computational methods to extract information from data. Neural network methods, generally regarded as forming the first wave of breakthrough in machine learning, became popular in the late 1980s, while kernel methods arrived in a second wave in the second half of the 1990s. This is the first single-authored textbook to give a unified treatment of machine learning methods and their applications in the environmental sciences.

Machine learning methods began to infiltrate the environmental sciences in the 1990s. Today, thanks to their powerful nonlinear modeling capability, they are no longer an exotic fringe species, as they are heavily used in satellite data processing, in general circulation models (GCM), in weather and climate prediction, air quality forecasting, analysis and modeling of environmental data, oceanographic and hydrological forecasting, ecological modeling, and in the monitoring of snow, ice and forests, etc.

This book presents machine learning methods and their applications in the environmental sciences (including satellite remote sensing, atmospheric science, climate science, oceanography, hydrology and ecology), written at a level suitable for beginning graduate students and advanced undergraduates. It is also valuable for researchers and practitioners in environmental sciences interested in applying these new methods to their own work.

Chapters 1-3, intended mainly as background material for students, cover the standard statistical methods used in environmental sciences. The machine learning methods of chapters 4-12 provide powerful nonlinear generalizations for many of these standard linear statistical methods. End-of-chapter review questions are included, allowing readers to develop their problem-solving skills and monitor their understanding of the material presented. An appendix lists websites available for downloading computer code and data sources. A resources website is available containing datasets for exercises, and additional material to keep the book completely up-to-date.

About the Author
WILLIAM W. HSIEH is a Professor in the Department of Earth and Ocean Sciences and in the Department of Physics and Astronomy, as well as Chair of the Atmospheric Science Programme, at the University of British Columbia. He is internationally known for his pioneering work in developing and applying machine learning methods in environmental sciences. He has published over 80 peer-reviewed journal publications covering areas of climate variability, machine learning, oceanography, atmospheric science and hydrology.

Machine learning methods in the environmental sciences 1st Table of contents:

1 Basic notions in classical data analysis

1.1 Expectation and mean

1.2 Variance and covariance

1.3 Correlation

1.3.1 Rank correlation

1.3.2 Autocorrelation

1.3.3 Correlation matrix

1.4 Regression

1.4.1 Linear regression

1.4.2 Relating regression to correlation

1.4.3 Partitioning the variance

1.4.4 Multiple linear regression

1.4.5 Perfect Prog and MOS

1.5 Bayes theorem

1.6 Discriminant functions and classification

1.7 Clustering

Exercises

2 Linear multivariate statistical analysis

2.1 Principal component analysis (PCA)

2.1.1 Geometric approach to PCA

2.1.2 Eigenvector approach to PCA

2.1.3 Real and complex data

2.1.4 Orthogonality relations

2.1.5 PCA of the tropical Pacific climate variability

2.1.6 Scaling the PCs and eigenvectors

2.1.7 Degeneracy of eigenvalues

2.1.8 A smaller covariance matrix

2.1.9 Temporal and spatial mean removal

2.1.10 Singular value decomposition

2.1.11 Missing data

2.1.12 Significance tests

2.2 Rotated PCA

2.3 PCA for vectors

2.4 Canonical correlation analysis (CCA)

2.4.1 CCA theory

2.4.2 Pre-filter with PCA

2.4.3 Singular value decomposition and maximum covariance analysis

Exercises

3 Basic time series analysis

3.1 Spectrum

3.1.1 Autospectrum

3.1.2 Cross-spectrum

3.2 Windows

3.3 Filters

3.4 Singular spectrum analysis

3.5 Multichannel singular spectrum analysis

3.6 Principal oscillation patterns

3.7 Spectral principal component analysis

Exercises

4 Feed-forward neural network models

4.1 McCulloch and Pitts model

4.2 Perceptrons

4.3 Multi-layer perceptrons (MLP)

4.4 Back-propagation

4.5 Hidden neurons

4.6 Radial basis functions (RBF)

4.7 Conditional probability distributions

4.7.1 Mixture models

Exercises

5 Nonlinear optimization

5.1 Gradient descent method

5.2 Conjugate gradient method

5.3 Quasi-Newton methods

5.4 Nonlinear least squares methods

5.5 Evolutionary computation and genetic algorithms

Exercises

6 Learning and generalization

6.1 Mean squared error and maximum likelihood

6.2 Objective functions and robustness

6.3 Variance and bias errors

6.4 Reserving data for validation

6.5 Regularization

6.6 Cross-validation

6.7 Bayesian neural networks (BNN)

6.7.1 Estimating the hyperparameters

6.7.2 Estimate of predictive uncertainty

6.8 Ensemble of models

6.9 Approaches to predictive uncertainty

6.10 Linearization from time-averaging

Exercises

7 Kernel methods

7.1 From neural networks to kernel methods

7.2 Primal and dual solutions for linear regression

7.3 Kernels

7.4 Kernel ridge regression

7.5 Advantages and disadvantages

7.6 The pre-image problem

Exercises

8 Nonlinear classification

8.1 Multi-layer perceptron classifier

8.1.1 Cross entropy error function

8.2 Multi-class classification

8.3 Bayesian neural network (BNN) classifier

8.4 Support vector machine (SVM) classifier

8.4.1 Linearly separable case

8.4.2 Linearly non-separable case

8.4.3 Nonlinear classification by SVM

8.4.4 Multi-class classification by SVM

8.5 Forecast verification

8.5.1 Skill scores

8.5.2 Multiple classes

8.5.3 Probabilistic forecasts

8.6 Unsupervised competitive learning

Exercises

9 Nonlinear regression

9.1 Support vector regression (SVR)

9.2 Classification and regression trees (CART)

9.3 Gaussian processes (GP)

9.3.1 Learning the hyperparameters

9.3.2 Other common kernels

9.4 Probabilistic forecast scores

Exercises

10 Nonlinear principal component analysis

10.1 Auto-associative NN for nonlinear PCA

10.1.1 Open curves

10.1.2 Application

10.1.3 Overfitting

10.1.4 Closed curves

10.2 Principal curves

10.3 Self-organizing maps (SOM)

10.4 Kernel principal component analysis

10.5 Nonlinear complex PCA

10.6 Nonlinear singular spectrum analysis

Exercises

11 Nonlinear canonical correlation analysis

11.1 MLP-based NLCCA model

11.1.1 Tropical Pacific climate variability

11.1.2 Atmospheric teleconnection

11.2 Robust NLCCA

11.2.1 Biweight midcorrelation

11.2.2 Inverse mapping

11.2.3 Prediction

Concluding remarks

Exercises

12 Applications in environmental sciences

12.1 Remote sensing

12.1.1 Visible light sensing

Ocean colour

Land cover

12.1.2 Infrared sensing

Clouds

Precipitation

Sea ice

Snow

12.1.3 Passive microwave sensing

12.1.4 Active microwave sensing

Altimeter

Scatterometer

Synthetic aperture radar

12.2 Oceanography

12.2.1 Sea level

12.2.2 Equation of state of sea water

12.2.3 Wind wave modelling

12.2.4 Ocean temperature and heat content

12.3 Atmospheric science

12.3.1 Hybrid coupled modelling of the tropical Pacific

12.3.2 Climate variability and climate change

Arctic Oscillation (AO)

Pacific?North American (PNA) teleconnection

Quasi-Biennial Oscillation (QBO)

Madden?Julian Oscillation (MJO)

Indian summer monsoon

Climate change

12.3.3 Radiation in atmospheric models

12.3.4 Post-processing and downscaling of numerical model output

Variance

Extrapolation

12.3.5 Severe weather forecasting

Tornado

Tropical cyclone

Hail

12.3.6 Air quality

12.4 Hydrology

12.5 Ecology

People also search for Machine learning methods in the environmental sciences 1st :

automated machine learning methods systems challenges

what are different machine learning methods

which of the following are machine learning methods

unsupervised machine learning methods

explain the different types of machine learning methods

Tags: William W Hsieh, Machine learning, environmental sciences