INF-3910-7 Computer Science Seminar: Computational Intelligence and its Applications - 10 ECTS
Admission requirements: Higher Education Entrance Requirement + Bachelor's degree in Computer Science or similar education. The Bachelor¿s degree must contain a specialization in Computer Science worth the equivalent of not less than 80 ECTS credits.
Application code: 9371 - Singular courses at master's level
Recommended prerequisites: Bachelor¿s in computer science or any engineering field (or similar). Programming skills. MAT-0001 / MAT-1001, MAT-1005, STA-0001 / STA-1001, MAT-1004.
Computational Intelligence (CI) is inspired from statistical, pattern recognition, neural network, machine learning, fuzzy logic, evolutionary computing, scientific visualization and other sources. This course covers basic CI techniques (details below), the use of a free software package WEKA, the use of a commercial software package MATLAB, and examples of practical applications of CI methods for data in technical, medical and bioinformatics domains.
The course addresses basic computational intelligence techniques for analyzing and designing approaches for different types of problems and applications of CI. It will also discuss emerging trends of computational intelligence from the international research front. Particularly, the following topics will be addressed:
- Overview of computational intelligence, types of adaptive systems, learning and applications;
- Data visualization, exploration, and analysis: Principal Component Analysis (PCA), Multidimensional Scaling (MDS), Self-Organized Mappings (SOM), parallel coordinates and other visualization algorithms;
- Theory: overview of statistical approaches to learning, bias-variance decomposition, expectation maximization algorithm, model selection, evaluation of results, ROC curves;
- Introduction to WEKA software packages, presentation of algorithms available in the package;
- Statistical algorithms for data analysis beyond classification: discriminant analysis - linear (LDA), Fisher (FDA), regularized (RDA), probabilistic data modelling, kernel methods;
- Neural Network for offline and online learning; Mamdani and Takegi-Sugeno-Kang models, Fuzzy inference systems, and convolutional neural networks;
- Evolutionary computation, Genetic Algorithm, Particle Swarm Optimization;
- Analyze and apply CI techniques for solving classification and regression tasks using Weka/Matlab
The focus is on using computational intelligence systems and techniques to analyze data and develop approaches for solution of different problems.
There will be 4 lab-sessions, one for WEKA (2.5 hours) and three for Matlab (2.5 hours each).
Knowledge - The student has:
- a foundation in computational intelligence systems and techniques;
- a conceptual understanding of the mathematical foundations of data analysis;
- a broad understanding of CI by introducing additional special topics into the curriculum.
Skills - The student can
- implement CI techniques on state-of-the art computer systems in solving real world problems;
- develop research and collaborative learning;
- work with different software packages for data analysis;
- accomplishing the tasks of the project;
- perform analyses using the CI systems built in the course;
- perform data analysis in practice, for academic or industrial data analysis tasks.
General competence - The student has
- aggregated understanding of different approaches of data analysis, emerging trends in CI with special focus on CS system aspects;
- become able to explore new CI concepts by practically implementing them on computer systems;
- an overview of emerging topics in CI.
The portfolio assessment consist of:
- Short assignment (individually assessed)
- Group report (one report per group of 2 or 3 students, all students of a group receive the same scores)
- Individual project report
- 4 hour written examination
Short assignment: There will be a short assignment for data analysis and visualization.
Group report: This component will be done in groups of 2 or 3 students. It deals with exposure of students to the emerging topics. Each group will be given few research papers to study and submit a report.
Individual project report: There is a project component in which each student can choose a dataset from a library of datasets, solve the assigned classification or regression tasks using methods of their choice learnt in the course and submit a report.
4 hours written examination: There is a written examination component, in which the students will be tested for the knowledge and understanding acquired during the course.
The exam parts are assessed as a whole and one combined grade is given.
Assessment scale: Letter grades A-E, F (fail)
Re-sit examination: A re-sit examination will not be given in this course.
Postponed examination: Students with valid reasons for absence will be offered a postponed examination for the module in question. If practically possible, the examination is arranged during the semester as soon as the reasons for absence have ceased.
Date for examinationWritten 12.12.2019
Short assignment hand out date 28.08.2019 hand in date 13.09.2019
Group report hand out date 17.09.2019 hand in date 08.10.2019
Individual report hand out date 15.10.2019 hand in date 10.11.2019
The date for the exam can be changed. The final date will be announced in the StudentWeb early in May and early in November.