The Swiss engineering master's degree

This course will provide an introductory review of the basic concepts of probability and statistics to understand probability distributions and to produce rigorous statistical analysis including estimation, hypothesis testing, and confidence intervals. Students will be introduced to the basic concepts of predictive modelling which by definition is the analysis of current and historical facts to make predictions about future events. Students will learn several techniques that account for many business and engineering applications of predictive modelling. These include regression techniques, time series models, and classification methods. Applicability and limitations of these methods will be illustrated in the light of data sets and analyses using the statistical software R or Python.

Please note: An MSE cursus may not contain both similar statistics modules FTP_AppStat and FTP_PredMod. Students can only choose one of these

modules.

Requisiti

Basic knowledge of statistics on the level of an introductory stochastics course. Linear algebra: matrix-vector calculations. Basic Calculus.
Familiarity and experience with programming, in particular with scripting languages like Matlab, Python or R. We will provide the students with a self-test to assess their prior knowledge in statistics and scripting.

Obiettivi di apprendimento

Students are able to analyze data by means of regression analysis. They are familiar with important statistical forecasting methods and are able to calculate, evaluate and interpret predictions. They are able to choose an appropriate statistical method for a regression, classification or time series problem. They are able to evaluate and compare statistical models.

Contenuti del modulo

Regression analysis: Simple linear regression with parameter estimation, graphical model validation, transformation of variables, confidence and prediction intervals for parameters. Multiple linear regression with parameter estimation, statistical tests and confidence intervals for parameters, variable selection, and regularization methods.

Classification: Concepts of classification, logistic regression, model evaluation metrics and cross-validation, boosting, model-agnostic feature importance analysis

Time series analysis: STL decomposition; ARMA, seasonal and non-seasonal ARIMA, Holt-Winters models with parameter estimation, confidence and prediction bands, autocorrelation, and model selection; anomaly detection; spectral analysis.  Use-cases in economics, finance, and engineering.

Metodologie di insegnamento e apprendimento

Lecture and practical work on computer with the statistics software R or Python.

Bibliografia

Lecture notes will be available in addition to recommended book chapters.