Introduction to Artificial Intelligence

1. komponenta

Lecture type	Total
Lectures	15
Seminar	15

* Load is given in academic hour (1 academic hour = 45 minutes)

Introduce the student to topics in artificial intelligence (deep learning), their foundations, applications and challenges.
Students will learn the principles of deep learning and how to independently develop basic AI systems and define the tasks, which are tackled with AI.
The student will be able to precisely define the relation between the individual and AI technologies and their role in society and recognize the security issues and potential abuse of AI, with a special emphasis on ethical problems arising from the applications of AI in security and threat prevention.
The student will learn how to build basic AI applications in Python (Anaconda), from Python fundamentals up to a complete multi-layered artificial neural network, which recognizes handwritten digits.

Learning outcomes at the level of the programme to which the course contributes:
Argue and rationally defend one's own philosophical views.
Conduct an interdisciplinary evaluation by taking into account different scientific methodologies and views on selected topics and problems.
Consider the justifiability of various theoretical and practical suggestions, by taking into account the specifics of various walks of life and work environments.
Develop critical thinking and structured formal thinking.

Expected learning outcomes at the level of the course:
1. Critically evaluate the role of AI and DL in our society
2. Describe convolutional neural networks
3. Evaluate the security issues and potential abuse of deep learning
4. Explore the various applications of AI and DL with regard to ethical issues of data collecting.

Course content:
1. General introduction to AI and its theoretical frameworks and limitations.
2. Introduction to machine learning and deep learning. Anaconda installation.
3. Diagramatic representation of the process of supervised learning. Basic operations in Python.
4. Measuring the efficiency of machine learning. Loops in Python.
5. Logistic regression. Definitions in Python.
6. Historical context of cybernetics and deep learning. List comprehension in Python.
7. Backpropagation. Parsing CSV with Python.
8. Multilayer neural networks. SKLearn in Python.
9. A visual introduction to convolutional neural networks. MNIST dataset. Loading MNIST with SKLearn.
10. Privacy of digital data. Gimp installation. Basic image processing in Gimp (greyscale, crop and resize).
11. Security applications of artificial neural networks. Multilayer artificial neural networks in SKLearn.
12. Preventive criminology and ethical questions connected with crime prevention. Development of a complete handwritten digit recognition system.
13. Deep learning and racism - from data imbalance to prejudice. Experiments with a complete handwritten digit recognition system.
14. Towards and epistemology of artificial intelligence: comparing AI to humans and cognitive processes. Keras basics and installation. A complete convolutional network for MNIST in Keras.
15. Final discussion

PSI (2980) - izborni TZP - Regular studij - Psychology

KRO (3001) - izborni TZP - Regular studij - Croatology
PSI (2980) - izborni TZP - Regular studij - Psychology

KRO (3001) - izborni TZP - Regular studij - Croatology
FIL (1826): Izborni kolegiji - Regular studij - Philosophy and Culture
SOC (2960) - izborni TZP - Regular studij - Sociology

KRO (3001) - izborni TZP - Regular studij - Croatology
FIL (1826): Izborni kolegiji - Regular studij - Philosophy and Culture
SOC (2960) - izborni TZP - Regular studij - Sociology

FIL (1826): Izborni kolegiji - Regular studij - Philosophy and Culture
SOC (2960) - izborni TZP - Regular studij - Sociology

FIL (1826): Izborni kolegiji - Regular studij - Philosophy and Culture