| CODE | SSA5010 | ||||||||||||
| TITLE | Algorithms for Astroinformatics | ||||||||||||
| UM LEVEL | 05 - Postgraduate Modular Diploma or Degree Course | ||||||||||||
| MQF LEVEL | 7 | ||||||||||||
| ECTS CREDITS | 5 | ||||||||||||
| DEPARTMENT | Institute of Space Sciences and Astronomy | ||||||||||||
| DESCRIPTION | This study-unit has three main components: 1. Introduction to Algorithms, Complexity, and Computation This component introduces students to core concepts such as computer algorithms, the various programming paradigms, computation, abstract data structures, sorting, recursion and iteration, and Big O Notation. The student is then introduced to data structures such as stacks, queues, lists, heaps, tables, trees, and graphs and their associated algorithms. 2. Image Processing Algorithms Imaging algorithms that are commonly applied to astrophysical datasets are put forward. Capturing processes by linear systems as well as techniques to minimize distortion effects introduced by the atmosphere, ionosphere and the telescope, are discussed. A review of different transformation basis functions and spectral analysis methods will be given so that students will be able to comprehend the complexities of current state-of-the-art procedures. 3. Machine Learning Algorithms This component introduces the main theoretical concepts in machine learning. This include PAC learning, Gold's Theorem, and Statistical Learning Theory. It then proceeds to introduce students to interdisciplinary topics that strongly link space research and computer science. The theoretical concepts covered during the lectures will be demonstrated in the tutorial sessions in which participants will work on real astrophysical data collected through past or ongoing optical and radio surveys. The digital world is currently facing a deluge of data captured from large arrays that form part of high resolution sensors. A lot of data mining efforts are currently focusing on novel ways to extract information from the massive bases which are being created. Through the Machine Learning topics covered in this course, the students will be made aware of supervised and unsupervised learning methods that are being used to infer new knowledge from collected data. Tools from statistical analysis that can be applied to obtain a measure of accuracy of the results, are also covered. Study-unit Aims: The main goal of this study-unit is to introduce students to the current state-of-the-art image processing and machine Learning techniques used in astrophysics. This study-unit also introduces the main concepts in algorithmics, computation, and complexity. Through the various tutorials and exercises, students will enhance their Matlab programming skills and be able to carry out the given assignment. This will involve processing and extraction of added value information from real data captured during recent sky surveys. To demonstrate their understanding of the project, students will be expected to give a class presentation. Students following this study-unit will enhance their problem solving skills and learn how to apply complex computational algorithms to a wide ranging set of tasks. This study-unit also intends to serve as a capacity building exercise and will allow participants to not only come up with computation solutions but also investigate why and when certain methods fail and how these can be improved. Learning Outcomes: 1. Knowledge & Understanding: Students completing this study-unit should be able to understand the algorithms, and the associated theoretical concepts, used in imaging and processing pipelines. Knowledge of different representation spaces including the spatial, frequency and wavelets domains as well as the processing required to convert signals from one basis function into another, is expected to be gained. Students should also be able to understand and apply Machine Learning techniques that will be covered during the course. These include Neural Networks, Genetic Algorithms and Fuzzy Logic. Knowledge of statistical methods to quantify the error or make conclusions on numerical results, is expected. 2. Skills: The aim of this study-unit is to provide students with the necessary skill set to process real data. Tutorials and demonstrations will be carried out using recent astrophysical data to allow participants to know what information is available and the corresponding limitations. Apart from understanding the theory behind computational methods, students will be able to design, implement and test complex programs in the Matlab environment. Acquired skills will include knowledge and extraction of data from online repositories as well as data visualization. Participants following this study-unit should also be able to perform in-depth analysis of the obtained results, indicate any sources of error and fine tune algorithms to achieve better results. Main Text/s and any supplementary readings: - Machine Learning by Tom M. Mitchell. - Astronomical Image and Data Analysis by Jean-Luc Starck and Fionn Murtagh. - Pattern Recognition and Machine Learning (Information Science and Statistics) by Christopher M. Bishop. - Introduction to Algorithms by Cormen, Leisserson, and Rivest. |
||||||||||||
| STUDY-UNIT TYPE | Lecture and Tutorial | ||||||||||||
| METHOD OF ASSESSMENT |
|
||||||||||||
| LECTURER/S | John M. Abela (Co-ord.) Adam Gauci Kristian Guillaumier |
||||||||||||
|
The University makes every effort to ensure that the published Courses Plans, Programmes of Study and Study-Unit information are complete and up-to-date at the time of publication. The University reserves the right to make changes in case errors are detected after publication.
The availability of optional units may be subject to timetabling constraints. Units not attracting a sufficient number of registrations may be withdrawn without notice. It should be noted that all the information in the description above applies to study-units available during the academic year 2023/4. It may be subject to change in subsequent years. |
|||||||||||||