Degree in Business Intelligence and Data Analytics

Lead the transformation of companies through the use and analysis of data.

Methods of decision analysis

Back to Syllabus

Description
A course on quantitative methods is offered in a wide variety of disciplines, from the social sciences to business to the natural sciences. The same statistical methods are applied across disciplines. Therefore, it should not be surprising that the tools you will learn to use in this course will benefit you in your future courses and careers regardless of whether your career interest is Finance, Accounting, Strategy, Management or Marketing. In this course, you will learn many interesting methods for making healthy decisions as managers of the future. The first part of this course focuses on the principles or rational decision-making supported by the theory of probability. The second part extends these technics to computer assisted decision-making in particular the method of Machine Learning (ML) by the use of Neural Networks (NN). By doing so, the student is provided with one of the main tools of the former paradigm of decision-making over big data. Basic programming skills will be required as the second part of the course aims to apply ML techniques in Python environment.
Type Subject
Tercer - Obligatoria
Semester
Second
Course
2
Credits
4.00

Titular Professors

Karl Erik Jonas Viklund
Professor
Previous Knowledge

Basic knowledge about statistics and probability.
Basic knowledge about calculus.

Objectives

Learning Outcomes of this subject are:
LO.01 - Know terminology, notation and methods from quantitative research, concretely those related to inference.
LO.02 - Able to analyse and summarize information from lectures and materials provided by the teacher.
LO.03 - Understand and be able to implement ML algorithms in Python.

Contents

These are the topics that will be covered during the course:
First part
1. Probability Review
2. Conditional probability and Bayes' Theorem
3. Decision Trees
4. Multicriteria Decision
Second part
5. Multivariable Function analysis
6. Perceptrons
7. Neural Networks

Methodology

Weekly teaching will consist of one lecturing session to explain basic concepts and group problem-solving in class to apply knowledge to practical situations. Programming sessions are for problem-solving and final project purposes.

Evaluation

Continuous assessment has the following evaluation structure:

Midterm - 30%: Statement, resolution
Final Project - 30%: Statement, resolution, and presentation
Individual assignments grade - 30%: 3 homework
Attendance, participation, and classwork - 10%: Class deliverables

To incorporate the practical case and group grades to the evaluation scheme, the average of the grade of individual assignments must be 4 or above.
The assessment of the practical case will be as follows:
1. Statement: originality and degree of application (10%)
2. Resolution of the exercise: (60%) - 20% each part
3. Conclusions (10%)
4. Explanation of the cases (20%)

RETAKE POLICY: Retake exam will be cumulative and the grade from the retake exam will count as the grade book grade. Maximum grade to pass the course in the retake is 6.0. Individual, group assignment and practical cases must be uploaded to campus virtual before retake exam.

Evaluation Criteria
Basic Bibliography

The recommended textbook is:
1) "Decision Analysis for Management Judgment". Paul Goodwin and George Wright. Wiley 2009.

Luckily, the most important references for machine learning are available online.
2) "Neural networks and deep learning". Michael Nielsen, available

3) "Practical Deep Learning for Coders - Practical Deep Learning (fast.ai)", J. Howard, S. Gugger.

For a more formal exposition of the topic and perspectives on advanced ML techniques:

4) "Deep Learning" I. Goodfellow, Y. Bengio, A Courville
https://www.deeplearningbook.org/

Additional Material