Data Visualization (Spring semester)

Schedule

  • Lectures and practice: Fridays 15:00-18:00 308 (MIF Didl.).
  • Lectures (Week I, VU Academic Calendar) 15:00-17:00 308 (MIF Didl.).

Laboratory Work

  • First exercise: Choose a multidimensional data set - from MSc thesis research, work, hobbies, or Internet (e.g. UCI Machine Learning Repository). Describe the data set: meaning, numbers and properties of objects (instances) and features (attributes, parameters).
  • Second exercise: Visualize multidimensional data set using at least three direct visualization methods. Matlab (Statistical Toolbox), Orange, Visulab, Xmdv, Pandas, or other software may be used.
  • Third exercise: Perform Principal Component Analysis on multidimensional data set. Determine, what percentage of the total variance is in each component, what percentage is within two principal components. Orange, Matlab, octave, or other software may be used.
  • Fourth exercise: Visualize multidimensional data set using Multidimensional Scaling. Investigate, how images are different when different initial solutions are specified. Orange, Matlab (drtoolbox, Sammon), SMACOF, Smooth 4.0, or other software and algorithms may be used.

Literature