Study Guide 2024-2027

Solid understanding of what deep learning is, when it’s applicable, and what its limitations are. Standard workflow for approaching and solving machine-learning problems and knowledge how to address commonly encountered issues. Ability to use Keras to tackle real-world problems ranging from computer vision to natural-language processing.

• Introduction: motivation, history of deep learning, deep learning today.
• Basics of model training.
• Gradient descent and backpropagation.
• Theoretical connection between maximum likelihood estimation and loss functions.
• Learning tasks (classification and regression) and output activations.
• Stochastic gradient descent, momentum and learning rate scheduling.
• Adaptive optimizers.
• Initialization and normalization.
• Regularization techniques.
• Basics of convolutional neural networks (CNNs).
• Advanced CNN architectures.
• Transfer learning.
• Sequential data and natural language processing.

Written exam or project work

Video lectures, voluntary homework, discussion on course piazza and weekly meetings (remotely or on campus).

Deep Learning Book (https://www.deeplearningbook.org/)

Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow, 3rd Edition (https://www.oreilly.com/library/view/hands-on-machine-learning/9781098125967/)

Video lectures and lecture notes, see https://users.utu.fi/knuutila/ for more information.

Good knowledge of Python programming and numpy is beneficial.

TKO_2115 Artificial Intelligence: Methods TKO_3107 Data structures and Algorithms

0-5

English

Advanced Studies

Computer Science

Csaba Raduly-Baka