1: \section*{Preface}
2: \addcontentsline{toc}{chapter}{Preface}
3: I have just gone through my email archive. The first exchange of messages with M. J. Cardoso, M. D., dates from the end of 2001.
4: It was Pedro Cardoso, her brother and my superior at INESC Porto, where I was a researcher and developer since 1999, who had put us in contact. She had just started her PhD and would probably need some assistance from someone skilled in software development and mathematics. Although I was already enrolled to start my own PhD in the beginnings of 2002, I accepted.
5:
6: Simultaneously, since the end of 2000, I had been working at INESC Porto for the MetaVision project. The MetaVision project proposed an innovative electronic production system to reduce the cost of film production and to allow more artistic flexibility in shooting and film editing. It also provided the enabling technology for the integration of real and virtual images at source quality for film production and in TV studios in the compressed domain.
7:
8: 2004 brought with it the end of the MetaVision project. That represented some free time that was exploited to fill some gaps detected in my mathematical background, by engaging in a masters in engineering mathematics. This master offers a solid formation in diverse areas of applied mathematics, divided in four main areas, comprising the analysis and processing of information.
9:
10: Coincidently, 2004 would also be the year of tight collaboration with M. J. Cardoso.
11: Her aim was to develop an objective measure for the overall cosmetic result of breast cancer conservative treatment.
12: When confronted with the problem, a machine learning approach (a topic that I was delving in the master's classes) emerged as the right move. A suggestion was made to predict the overall cosmetic result from a few simple measures taken from the patient. I knew already some tools to tackle the problem but only superficially. That led me to select the \textsl{automatic classification and pattern recognition}, lectured by Professor Joaquim F. Pinto da Costa, as one of the modules to attend.
13:
14: The application of some of the state of the art methods for ordinal data, to the problem at hand, sparkled the interest on this specific topic of classification. What I learned, the breakthroughs that were accomplished, is what I would like to share with you.
15:
16: \vspace{0.4cm}
17: This work became possible due to the support of Professor Joaquim F. Pinto Costa, my supervisor. I also discussed ideas presented in this thesis with L. Gustavo Martins, Lu\'is F. Teixeira and M. Carmo Sousa. They also read the manuscript, providing important feedback. I would like to express my deep gratitude to all of them.
18:
19: \vspace{0.4cm}
20: \emph{Jaime dos Santos Cardoso}\\
21: \emph{September 2005}