| CODE | CCE5221 | |||||||||
| TITLE | Pattern Recognition and Machine Learning | |||||||||
| UM LEVEL | 05 - Postgraduate Modular Diploma or Degree Course | |||||||||
| MQF LEVEL | Not Applicable | |||||||||
| ECTS CREDITS | 6 | |||||||||
| DEPARTMENT | Communications and Computer Engineering | |||||||||
| DESCRIPTION | This study-unit describes, derives and evaluates various Pattern Recognition techniques that can be later applied in various domains. Both the theoretical and practical aspects of the pattern recognition techniques are dealt with. At the end of the study-unit the student will be able to select, develop and apply machine learning algorithms to solve particular problems. The study-unit is organized in three parts. The first part makes use of linear, polynomial and logistic regression models to define the pattern recognition problem and to describe how a pattern recognition system is designed and built. Common considerations in pattern recognition, such as over-fitting, under-fitting, regularization, model order and dimensionality are introduced in this part. A review of optimisation techniques, appropriate cost functions and the respective analytical solutions and numerical solutions are covered for these techniques. The second part of this study-unit introduces higher order features in view of non-linear models and motivates the concept using Neural Networks, followed by kernel methods and support vector machines. The clustering problem is introduced with the k-means algorithm and principle component analysis is used to motivate dimensionality reduction. This part is concluded with practical aspects when building a machine learning system; learning curves, error analysis, and precision and recall. The third part of this study-unit covers density estimation, probabilistic and Bayesian methods. It therefore starts with a review of statistics and probability. Maximum likelihood is formalized in Gaussian mixture models, and the regression problem is viewed from a probabilistic perspective, frequentist and Bayesian approach, to gain insight in the minimization of error functions and regularization. The Expected Maximisation algorithm is motivated in GMM models and is derived from both a probabilistic and aBayesian perspective. The final part of the study-unit covers Bayesian classifiers (MAP and Naive Bayes) and ends with graphical probabilistic models; Bayes Networks and Hidden Markov Models. Learning Outcomes: 1. Knowledge & Understanding: By the end of the study-unit the student will be able to: - Comprehend the concepts of Pattern Recognition and the Features to be used for the classification; - Comprehend various Machine Learning algorithms available including Neural Networks, Support Vector Machines and Hidden Markov Models. 2. Skills: By the end of the study-unit the student will be able to: - Implement statistical representation of a problem; - Design features and classifiers to solve a particular problem. Main Text/s and any supplementary readings: - Kevin P. Murphy, Machine Learning; A probabilistic Perspective, MIT press, 2012. - Christopher M. Bishop, Pattern Recognition and Machine Learning, Springer, 2006. - Tom Mitchell, Machine learning, McGraw Hill, 1997. - Richard O. Duda, Peter E/ Hart and David G. Stork, Pattern Recognition, Second Edition, 2001. - Simon Haykin, Neural Networks and Learning Machines, 3rd edition, Pearson, 2009. |
|||||||||
| RULES/CONDITIONS | Before TAKING THIS UNIT YOU ARE ADVISED TO TAKE CCE2501 AND TAKE MAT1801 AND TAKE SOR1201 | |||||||||
| STUDY-UNIT TYPE | Lecture | |||||||||
| METHOD OF ASSESSMENT |
|
|||||||||
| LECTURER/S | Michael Camilleri Adrian F. Muscat (Co-ord.) |
|||||||||
|
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. |
||||||||||