Code

ΜΔΑ-289

Semester

2nd

ECTS

7,5

E-Services

Lefkippos

Category

Obligatory

Instructors

I. Maglogiannis
Α. Prentza

Objective

The aim of the course is to introduce advanced machine learning and artificial intelligence methodologies related to deep learning, performance evaluation and the combined use of basic algorithms, and the preparation and processing of available data for their more efficient use. Expected learning outcomes include a thorough understanding of the performance of deep learning methods, the ability to use them in combination to solve challenging problems, and the ability to analyze data to pre-process it and combine it with the appropriate methodology.

After successfully completing the course, students will be able to:

  • explain fundamental concepts of artificial intelligence
  • choose an algorithm for solving artificial intelligence problems
  • evaluate the usefulness and weaknesses of alternative algorithms and techniques
  • model problems as search, constraint solving and logic problems
  • understand deep learning architectures
  • design and implement deep learning systems
  • evaluate the appropriateness of implementing deep learning systems

Learning outcomes

  • Search for, analysis and synthesis of data and information, with the use of the necessary technology
  • Adapting to new situations
  • Decision-making
  • Working independently
  • Team work
  • Production of new research ideas
  • Project planning and management
  • Criticism and self-criticism
  • Production of free, creative and inductive thinking

Syllabus

  • Introduction

    Introduction to artificial intelligence and machine learning, problem categories, supervised learning, unsupervised learning, reinforcement learning, examples of applications.

  • Neural networks I

    Introduction to neural networks, neural network models and architectures, perceptron, linear and non-linear separability, multilayer perceptron, neural network training algorithms.

  • Neural networks II

    Performance evaluation of neural networks, generalizability, neural network development applications, case study.

  • Clustering I

    Definitions, clustering classes, distance functions, similarity functions, partitional clustering, k-means algorithm.

  • Clustering II

    Hierarchical clustering, evaluation and validity of clustering, applications of clustering, case study.

  • Deep learning and convolutional neural networks I

    Introduction to deep learning, concatenation and clustering, deep learning architectures, training deep neural networks.

  • Deep learning and convolutional neural networks II

    Recurrent neural networks, genetic models, detection and segmentation, visualization and understanding, transfer learning.

  • Deep Learning Lab

    Examples of deep learning, Recognition with pre-embedded networks, learning transfer, training and evaluation.

  • Multidimensional data processing

    Multidimensional vision, Feature extraction, Recognition, Classification, Video analysis.

  • Machine learning in biomedical data

    Biomedical data representation, Knowledge extraction, Event and anomaly detection in medical history, machine learning for diagnosis and health strategies.

Bibliography

  • Machine Learning. McGraw-Hill (International Edition), 1997.
  • Pattern Recognition and Machine Learning. Springer, 2007.
  • Python for Data Analysis. O’Reilly, 2012.
  • Python Machine Learning. Packt Publishing, 2015.
  • Witten, E. Frank, M. A. Hall. Data Mining: Practical Machine Learning Tools and Techniques. Morgan Kaufmann, 3rd edition, 2011.
  • Ian Goodfellow and Yoshua Bengio and Aaron Courville, Deep Learning, MIT Press, http://www.deeplearningbook.org, 2016
  • Journal of Machine Learning Research (www.jmlr.org)
  • Machine Learning (Springer)
  • Pattern Recognition (Elsevier)
  • Journal of Artificial Intelligence Research (www.jair.org)