72534 - Numerical Methods of Statistics

Learning outcomes

At the end of the course the student has acquired: -the knowledge of software tools for the data analysis - the definition and the main characteristics of continuous and discrete distributions with their moments -The Maximum Likelihood method for parameters estimation - some numerical methods for linear regression and its application in economy - the main concepts of Statistical Learning for regression and classification

Course contents

Definitions and examples on probability. Discrete and continuous distributions. Descriptive statistics. Points and interval estimations. Hypothesis test. Least squares data approximation: linear regression, polynomial functions of higher order and nonlinear least squares. Numerical methods for the solution of the discrete linear least squares problem.

Simulations and programming environment R. Principal functions for graphics and data analysis. Guided exercises on examples with simulated and real data.

Introduction to statistical learning. Classification task.

Readings/Bibliography

j. Unpingco, Python for probability, statistics and machine learning, Springer

Teaching methods

Lessons and guided exercises on laptop.The exercises consist in small programs in the R

In relation to the kind of activities and didactical methods adopted, the attendance to this class will require the preliminary participation of all the students to the Modules 1 and 2 of the Safety rules on study places, which can ne followed remotely in e-learning via the following link: [https://elearning-sicurezza.unibo.it/environment on the main topics dealed with in the lessons.

Assessment methods

written exam with multiple choice quiz and oral discussion on an assigned project.

Teaching tools

Example program files given by the teacher

Office hours

See the website of Elena Loli Piccolomini