Content, progress and pedagogy of the module

This module provides an applied introduction to deep learning, progressing from foundational neural network architectures (ANNs, CNNs, RNNs, and LSTMs) to advanced approaches such as Transformer models, Generative Pre-trained Transformers (GPTs), Graph Neural Networks (GNNs), and Graph Transformers. Along the way, students gain practical insights into training paradigms, attention mechanisms, autonomous agent systems, and graph-based attention models, with applications in areas including customer behavior prediction, market trend analysis, sentiment analysis, recommendation systems, fraud detection, and supply chain optimization. 

Teaching combines lectures, exercises, and case-driven discussions with a strong emphasis on hands-on learning. Students work in groups on four mini-projects, building interactive demos and a portfolio of applied AI solutions. By the end of the module, participants will have acquired both theoretical understanding and practical competence to design, implement, and critically evaluate deep learning models for real-world business challenges. 

Learning objectives

Knowledge

By the end of the course, students will have knowledge of: 

• Central concepts in deep learning and the key components of artificial neural networks. 

• Major supervised and unsupervised architectures, including recent developments such as Transformers, GPT models, Graph Neural Networks, and Graph Transformers. 

• Ethical and societal challenges arising from AI adoption, including issues of bias, fairness, transparency, and sustainability 

Skills

By the end of the course, students will be able to: 

• Select and prepare diverse datasets for deep learning tasks, including structured, unstructured, and graph-based data. 

• Design and implement architectures such as ANNs, CNNs, RNNs, Transformers, GPTs, and Graph Transformers, including the use of large pretrained models and fine-tuning strategies. 

• Train, tune, and evaluate models for predictive tasks, optimization, and deployment in business-relevant scenarios. 

• Build applied projects and interactive demos that showcase deep learning techniques in practice. 

• Describe and compare the aforementioned supervised and unsupervised architectures. 

• Recognize and articulate the ethical and societal challenges related to AI adoption. 

Competences

By the end of the course, students will be capable of: 

• Applying deep learning and AI techniques to solve business challenges across domains such as marketing, finance, operations, and decision support. 

• Making informed decisions about the selection of algorithms, including when simpler methods may be more appropriate than deep learning. 

• Identifying cases that require particular attention concerning the ethical and social consequences of deep learning and AI applications. 

Type of instruction

For information see § 17.

Exam

Prerequisite for enrollment for the exam

• A prerequisite for participating in the exam is that the student has handed in the required written material (portfolio assignments and project documentation).

Exams