In this curricular unit, we intend to identify the fundamental notions that characterize the structure of a network, to be able to analyze concrete networks of the real world, calculating appropriate metrics, communities, etc., and extracting information / insights from the data. It is also intended to study some efficient algorithms used in the analysis of complex networks, to understand some network behavioral models and to simulate a propagation process in a network.
This curricular unit will study structures of large networks and graphs, addressing issues such as connectivity, centrality and communities; prediction of processes that may occur in large graphs; and prediction of the growth of large real graphs.
In this curricular unit the student will make contact with Bayesian models for the analysis of very complex data structures. The principles governing Bayesian inference, techniques for incorporating existing a priori knowledge and the corresponding uncertainty into a probability distribution, as well as hierarchical modelling to represent and analyse complex systems, will be exposed.
This course unit aims at giving the student a general perspective of large-scale data processing. The main existing solutions for the storage and distributed processing of data are reviewed. Regarding the processing tools studied, special emphasis is given to the supported programming models, from a theoretical-practical perspective, based on the study of examples and resolution of exercises. In the practical component of the UC, a project is developed that addresses a non-trivial problem, using the Hadoop ecosystem suite of tools, where the Apache Spark system stands out.
This course unit deals with concepts related to inference in large data. Particular attention will be given to statistical methods, sampling-resampling and simulation techniques.
This course focusses on the areas of Data Analysis and Text Mining. The Data Analysis module covers topics of pre-processing, dimensionality reduction, partitioning clustering algorithms and spectral clustering (crisp vs. fuzzy), validation and interpretation of clustering. The Text Mining module focuses on the extraction of Relevant Information, the symbolic and statistical analysis of the texts, the classification of documents, the automatic construction of its descriptors and the distribution of words in the context of Big Data.
This course will deal with issues related to decision making, game theory, negotiation models and strategic interaction between the players in the decision-making process. Different concepts and models of risk will also be studied.
This course is intended to motivate students for entrepreneurship and the need for technological innovation. It covers a list of topics and tools that are important for new venture creation as well as for the development of creative initiatives within existing enterprises.
How to research medical information to help decide a patient's diagnosis? How to search for all news related to a natural disaster? This course addresses these technical issues by providing students with an understanding of the design and implementation of search engines. Students will master fundamental concepts of Information Retrieval: information representation, indexing, query and classification by relevance.
This curricular unit presents the theoretical and practical bases for the conception, use and evaluation of modern systems for interactive visualization of data. In addition to the fundamentals of Visualization, the main techniques of visualization of multivariate data, geo-spatial data, and time-dependent data are presented.
Most digital information is not structured, requiring a laborious process to extract knowledge of video, sound, images or free text using traditional analytical techniques. This curricular unit presents recent deep learning techniques that allow direct dealing with this data, including deep neural networks, convolution and recurrent, deep generator models, pre-training and practical examples of application with the Tensorflow library.
This course introduces the student to linear optimization. Topics related to modelling, solving and analysing solutions of linear, linear and integer linear programming problems with multiple objectives will be addressed.
The increasing abundance of information makes automatic knowledge extraction increasingly important. This curricular unit covers the fundamentals of automated learning, including different classification and regression models, and the over-adjustment problem in supervised learning, and agglomeration and attribute extraction algorithms in unsupervised learning.
This curricular unit focusses on the analysis of multivariate data. In particular, aspects related to the main statistical distributions, inference for means vectors and covariance matrices, techniques for data classification and dimensionality reduction, and the application of different techniques and statistical methods in the analysis of (large volumes) data multivariate.
This course will focus on modelling and solving non-linear optimisation problems. It will cover topics that will allow the student to distinguish easy problems from those of difficult resolution and to use the main methods of nonlinear optimization with the knowledge of their numerical weaknesses. It will also analyse possible resolution approaches for optimization problems of this nature.
In this course students are exposed to various case studies, in which several institutions (companies, public institutions, research centres, etc.) obtain and / or manage large volumes of data and want to take advantage from the knowledge and useful information extracted from these data. Students should examine these case studies, presented by representatives of these institutions, in order to analyse the possibilities of extraction and use of this knowledge to improve the decisions to be taken.
In recent years there has been a huge increase in the amount of information that is continuously generated (for example, by financial systems, telecommunications networks and public services or sensor networks). This course aims to study the fundamentals, languages and systems for the construction of applications that process data streams, from generic real-time distributed stream processing systems to structured data models to deal with streams.