Pipelines for Smart Process Analytics

Recommended Prerequisites

Not applicable.

Teaching Methods

The teaching methodology is based on the presentation of contents and discussion of practical cases in an interactive way with students.

Small projects carried out by the students and their presentation and discussion conclude the learning and evaluation cycle.

Learning Outcomes

Acquisition of soft skills relevant for the design of data analysis strategies to handle complex data, depending on their structure, how they were acquired, objectives to be achieved and resources available.

Trainees will have contact with the main phases in the data analysis process, common difficulties, and how to get around them.

Work Placement(s)

No

Syllabus

1. Types of data.

2. Active and passive data collection

3. Pipeline for data analysis

i. Collectio

ii. Integration

iii. Cleaning, missing data imputation

iv. Pre-processing and feature engineering

v. Exploratory data analysis

vi. Model building

vii. Model validation

viii. Pilot solution

ix. Model maintenance plan

4. Analysis of the Quality of Information generated in the study (Info-Q)

6. Analysis of Case Studies.

Head Lecturer(s)

Marco Paulo Seabra dos Reis

Bibliography

• Bart Baesens, Analytics in a Big Data World, Wiley, 2014
• Bernard Marr, Big Data, Wiley, 2015

Dias, T., R. Oliveira, P.M. Saraiva, M.S. Reis, Predictive Analytics in the Petrochemical Industry: Research Octane Number (RON) forecasting and analysis in an Industrial Catalytic Reforming Unit. Computers & Chemical Engineering. 139 (2020), Article 106912. DOI: 10.1016/j.compchemeng.2020.106912