ASE-4046 Optimisation and Data Analysis, 5 cr

Additional information

Acceptable for postgraduate studies only if grade is at least 3.
Suitable for postgraduate studies.

Person responsible

Robert Piche

Lessons

Learning Outcomes

A hands-on introduction to optimisation and statistical modelling. %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Period 3: Optimisation %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 1. Computer arithmetic - floating point numbers - FP arithmetic 2. Linear programming - examples: production planning, transportation allocation, diet planning - graphical solution - Matlab LINPROG - ill-posed LP problems 3. Curve fitting - fitting a line and fitting a polynomial - data linearisation transformations - robust fit with linear programming 4. Nonlinear least squares - examples: positioning, curve fitting, feedback controller design - Gauss-Newton method - LSQNONLIN - ill-conditioned problems 5. Nonlinear optimisation - unconstrained, FMINUNC - Lagrange multipliers - quadratic cost with linear equality constraints - FMINCON 6. Multiobjective optimisation - Pareto optimality - weighted sum method - goal attainment, FGOALATTAIN - examples: shopping, controller design %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Period 4: Statistical data analysis %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 7. Visualising data - histogram, kernel PDF, CDF - medians, quantiles, box plots - data presentation do's and don'ts 8. Inference of categories - frequency diagram - Bayes formula - Bayesian nets, AISPACE software 9. Inference of Bernoulli parameter - binomial sampling model - posterior distribution and predictive distribution - updating: using prior information, sequential learning - parameter difference via Monte Carlo 10. Inference of Gaussian mean - Gaussian sampling model, normal QQ plot - posterior distribution & predictive distribution - updating: using prior information, sequential learning 11. Multiple linear regression - examples: fitting a line, fitting a polynomial - posterior distribution & predictive distribution - goodness of fit - updating: sequential learning 12. Kalman filter - state space (hidden Markov) model - Bayes filter - Kalman filter, steady-state KF, extended KF - examples: channel estimation, target tracking

Prerequisites

Additional information about prerequisites
Multivariate calculus, probability, and basic Matlab programming skills are mandatory prerequisites.

Correspondence of content

There is no equivalence with any other courses