MACHINE LEARNING AND DATA SCIENCE

Single discipline educational activity

Course Sheet

Course code:

MALDAS22

Academic Year of enrolment:

2022

Year of supply:

2022/2023

Type of Course:

Elective Course

Degree:

Course of 2nd Cycle Degree in DIGITAL MARKETING

Disciplinary Sector:

Data Processing Systems

Language:

Italian

Location:

PESCARA

Credits:

Programme Year:

Professor and Collaborators:

MORBIDONI Christian

Cycle:

Third four months

Hours of classroom activity:

Individual study hours:

114

Prerequisites:

Knowledge of the contents of the Statistics and Data Analytics courses is strongly recommended.

Objectives

Specific training objectives of the course are the following:
It distinguishes the characteristics and the different phases of the main Data Science project management models and their phases.
It distinguishes the different types of data in relation to the quantities to be represented and the operations supported
It composes basic workflows in KNIME, selecting and appropriately interconnecting the nodes.
It distinguishes the main data preparation operations and their purposes.
Compose data preparation workflows in KNIME using the appropriate nodes to manipulate and process the data
It distinguishes the characteristics and purposes of the different types of supervised and unsupervised Machine Learning problems.
It distinguishes the issues and stages of the evaluation process of a machine learning system.
Implement KNIME workflow to perform multiple regression analysis and result evaluation
It distinguishes the characteristics of the main machine learning approaches applied to Classification problems.
It defines the main metrics for evaluating a classification system.
Implement classification workflow in KNIME
It distinguishes the characteristics of the main machine learning approaches applied to Clustering problems.
Defines the main metrics for evaluating a clustering analysis.
Implement clustering workflow in KNIME

Contents

The course aims to provide the basic knowledge and practical skills to implement marketing data analysis processes based on Machine Learning techniques, through the use of Open Source and free tools.
The course consists of a theoretical part in which the basic concepts related to Data Science and Machine Learning algorithms and methods (Regression, Supervised Classification and clustering) are introduced and a practical one in which the concepts will be applied through the open source platform KNIME (e.g. for segmentation, buy / churn prediction, sentiment analysis).

Extended Syllabus

Module 1: Data Science and Machine Learning
- Introduction to Data Science: Definition, basic concepts.
- The phases of a Data Science project and process management models
- Data Understanding: Sources and types of data
- Introduction to Machine Learning: Definitions and characterization of the main problems and methods.
- Practical introduction to the KNIME platform and first analysis workflow
Module 2: Data preparation / visualization
- Data preparation: problems and practical solutions in KNIME, case studies and practical exercises
- Manipulation, aggregation and visualization of data in KNIME, case studies and practical exercises
Module 3: Modeling
Linear and logistic regression: recalls.
Regression workflow in KNIME: case study and practical exercise.
Classification algorithms and evaluation metrics.
Classification workflow in KNIME: case study and practical exercise.
Clustering algorithms: recalls
Segmentation workflow via clustering in KNIME: case study and practical exercise.

Recommended Bibliography

Study material provided by the teacher during the lessons.

Methods of Provision

Conventional

Teaching Methods

The course alternates theoretical lectures and classroom exercises in which students will be able to apply the methods studied in practice through the tool adopted (KNIME).

The teaching includes 36 hours of lessons.
The organization of the course is based on "Listening", an experimental model for the continuous improvement of the quality of university teaching. The course will be organised in modules whose lessons will mainly focus on the use of the following teaching methods: lectures, demonstrations of practical cases, exercises.

Evaluation methods

Verification of learning:

The final exam consists of a practical computer test with KNIME software and an oral interview.

It will be possible to agree on project work of a practical nature, which deepen the topics covered in class, to be discussed during the oral interview.

Università degli Studi
"G. d'Annunzio"

Chieti - Pescara

Aree

Error message