Advanced Artificial Intelligence; Machine Learning and Deep Learning in Practice

Educational goals

• Understand the basic principles behind machine learning.
• Learn how to build and evaluate machine learning models.
• Familiarize yourself with the practical applications of machine learning.
• Understand the basic concepts of neural networks and deep learning.
• Explore the use of neural networks in real-world applications.
• Understand how to train and evaluate deep learning models.

 Learning Outcomes

• Describe basic machine learning concepts and how to apply them.
• Explain model evaluation without the need for programming.
• Identify appropriate situations for supervised and unsupervised learning.
• Describe the structure of neural networks.
• Familiarize yourself with key applications of deep learning (images and text).
• Understanding the challenges of training deep learning models.

📘 General themes

1) Foundations and concepts

• What is data? Structured vs. unstructured.
• The impact of data biases on AI output.
• Explore public datasets (health, financial…)

2) Types of models

• Classification, regression, clustering – when do we use each type?
• Visually compare model outputs using IBM Watson tools.

3) No-Code Tools and AutoML

• Utilize IBM AutoAI and AutoML platforms.
• Build models without programming and design an AI-based solution.

4) Deep learning: Concepts

• What makes a model “deep”? Layers of neural networks.
• The differences between deep learning and traditional machine learning.
• Application case: Image recognition (e.g., medical imaging).

5) Neural Networks Clearly

• Input, hidden layers, output layer.
• Visualization of neural networks (Teachable Machine and others).
• Explore pre-trained IBM Watson models and interpret the outputs.

6) Deep Learning Applications

• CNNs: How do you “see” images?
• Image classification and face recognition + ethical dimensions.

7) Natural Language Processing (NLP)

• How do neural networks process text?
• Sentiment analysis with IBM Watson tools and examples from real data.

8) Evaluation and projects

• Conceptual test + group demonstration of deep learning applications in a selected industry.
• Applied project: Building a classification model on IBM AutoAI to predict customer type (new/return) based on marketing data.