Advanced Data Analysis Techniques
CVEN 6833

Fall 2018
Instructor: Prof. R. Balaji
Office Location: ECOT-444
Phone: (303)-492-5968
E-mail: balajir@colorado.edu
Lectures: Tue. and Th. 10:30 - 11:45 in SEEC N126

Course Material & Resources
R-Session/Homeworks
Relevant Research Papers

Information

Useful Books 
Linear Regression
Generalized Linear Models (GLM)
Reference: Prof. G. Rodriguez (2007) Notes
GLM - Quick tutorial

Nonparametric (Local) Regression

Kernel Density methods
Spatial Modeling - Kriging

Smoothing/Clustering/Trees/Data Mining

Self Organizing Maps (SOMs)

Cluster Analysis (Dr. McCreight)

Multivariate Methods

Time Series Analysis

Hidden Markov Models

Bayesian Modeling - BDLM & Hierarchical
Bayes resources - notes, presentations, etc.

Extreme Value Distribution Resources

Copula

Bayesian Dynamical Regression BDLM

Singular Spectrum Analysis

Spectral Analysis

R-links
Data Files for R-practice:
 Leeferry-mon-data.txt – Lee’s Ferry Data
Note: col1 = year, col2 = Jan… col13 = Dec
Units: acre-feet; divide by 10^6 to get MAF (Million Acre-Feet) or;
 (X 0.0004691 to get cubic meters per second (i.e., cms))
 
Leeferry-annual-maximum.txt
 aismr.txt – All Indian monthly rainfall total (in mm)
 nino3-index.txt – Nino3 Index of ENSO
annual-peak-clark-fork.txt – Annual Flood Data at Clark Fork River,
 (cubic feet / second)


Homeworks
R-Session Linear Regression
R-Session GLM/Local Polynomials
  •  1 and 2 variable data sets
  • Fit a best local polynomial model (using GCV)
  • Perform model diagnostics
  • Estimate the function at several locations
  • Plot/Map the estimates and error variance

R-JAGS Manual

R-Session Bayes
  •  Spatial Bayesian Hierarchical Modeling
  • Code/commands for results in Verdin et al. (2015)
  • STAN examples
  • spBayes - R Package

R-Session - Multivariate Methods 

  •  PCA
  • SVD
  • CCA
  • Multivariate Forecasting
  • SST and PDSI data (examples)

R-Session - Time Series Simulation/Projection

  • Parametric ARMA
  • Parametric Seasonal AR
  • Nonparametric AR
    • Nonparametric seasonal AR
  • Hidden Markov Model
  • Markov Chain + Resampling
  • Spectral Analysis
  • Mutual Information (MI)

Nonparametric Time Series Modeling

Local Polynomial Applications
GLMs GAM & Bayesian Dynamic Regression

Hierarchical Modeling - Kriging

CART
Extremes Clustering

Multivariate Modeling and Forecasting