Course Description
The philosophical and theoretical basis of informatics, with applications in civil engineering, environmental engineering, and the environmental sciences. Readily available software will be used throughout the course. Specific applications will depend on the needs of the students in the course.
Athena Title
Informatics Engr Environ Sci
Semester Course Offered
Offered fall
Grading System
A - F (Traditional)
Course Objectives
To develop an understanding of the philosophical and theoretical basis for informatics To develop programming skills needed to perform informatic techniques using commonly available software To apply informatic techniques to civil engineering, environmental engineering, and environmental sciences data To develop an ability to read, comprehend, and contribute to the literature in environmental engineering and sciences using informatic techniques.
Topical Outline
Philosophical basis for machine learning Theoretical basis for machine learning Review of classical regression analysis Regression analysis in the informatic context Application: global temperature trend Application: hydrologic flows Logistic regression Application: water quality Linear discriminant analysis Application: fault diagnosis: a case study in automotive industry Cross validation and bootstrapping Linear model selection and regularization Polynomial regression, step functions, basis functions, regression splines, smoothing splines, local regression, generalized additive models Application: smoothing of environmental data sets Tree-based models Application: A tree-based model: evaluating the thermal comfort of horses Support Vector Machines Application: renewable energy problems: the wind speed prediction from measurements in neighbor stations and the wind speed reconstruction using synopticÂ-pressure data. Unsupervised learning Application: temporal patterns in continental U.S. climate patterns Neural networks Application: use of neural networks to predict air pollution Labs Introduction to R Linear regression in R (libraries, SLR, MLR, interaction terms, writing functions) Logistic regression, LDA, QDA, and KNN Validation (validation set, leave-one-out, k-fold cross- validation, cross-validation on classification problems) and bootstrap Subset selection methods (best subset, forward and backward stepwise selection, choosing among models) Non-linear modeling (polynomial regression and step functions, splines, GAMs) Decision trees (fitting classification trees, fitting regression trees, bagging and random forest, boosting) Support vector machines (support vector classifier, support vector machine, ROC curves, SVM with multiple classes) Principal component analysis and cluster analysis Neural networks