Multi-dimensional classification with super-classes

The multi-dimensional classification problem is a generalisation of the recently-popularised task of multi-label classification, where each data instance is associated with multiple class variables. There has been relatively little research carried out specific to multi-dimensional classification and, although one of the core goals is similar (modelling dependencies among classes), there are important differences; namely a higher number of possible classifications. In this paper we present method for multi-dimensional classification, drawing from the most relevant multi-label research, and combining it with important novel developments. Using a fast method to model the conditional dependence between class variables, we form super-class partitions and use them to build multi-dimensional learners, learning each super-class as an ordinary class, and thus explicitly modelling class dependencies. Additionally, we present a mechanism to deal with the many class values inherent to super-classes, and thus make learning efficient. To investigate the effectiveness of this approach we carry out an empirical evaluation on a range of multi-dimensional datasets, under different evaluation metrics, and in comparison with high-performing existing multi-dimensional approaches from the literature. Analysis of results shows that our approach offers important performance gains over competing methods, while also exhibiting tractable running time.

Inteligencia artificial para la predicción y control del acabado superficial en procesos de fresado a alta velocidad

El objetivo principal de esta tesis doctoral es profundizar en el análisis y diseño de un sistema inteligente para la predicción y control del acabado superficial en un proceso de fresado a alta velocidad, basado fundamentalmente en clasificadores Bayesianos, con el prop´osito de desarrollar una metodolog´ıa que facilite el diseño de este tipo de sistemas. El sistema, cuyo propósito es posibilitar la predicción y control de la rugosidad superficial, se compone de un modelo aprendido a partir de datos experimentales con redes Bayesianas, que ayudar´a a comprender los procesos dinámicos involucrados en el mecanizado y las interacciones entre las variables relevantes. Dado que las redes neuronales artificiales son modelos ampliamente utilizados en procesos de corte de materiales, también se incluye un modelo para fresado usándolas, donde se introdujo la geometría y la dureza del material como variables novedosas hasta ahora no estudiadas en este contexto. Por lo tanto, una importante contribución en esta tesis son estos dos modelos para la predicción de la rugosidad superficial, que se comparan con respecto a diferentes aspectos: la influencia de las nuevas variables, los indicadores de evaluación del desempeño, interpretabilidad. Uno de los principales problemas en la modelización con clasificadores Bayesianos es la comprensión de las enormes tablas de probabilidad a posteriori producidas. Introducimos un m´etodo de explicación que genera un conjunto de reglas obtenidas de árboles de decisión. Estos árboles son inducidos a partir de un conjunto de datos simulados generados de las probabilidades a posteriori de la variable clase, calculadas con la red Bayesiana aprendida a partir de un conjunto de datos de entrenamiento. Por último, contribuimos en el campo multiobjetivo en el caso de que algunos de los objetivos no se puedan cuantificar en números reales, sino como funciones en intervalo de valores. Esto ocurre a menudo en aplicaciones de aprendizaje automático, especialmente las basadas en clasificación supervisada. En concreto, se extienden las ideas de dominancia y frontera de Pareto a esta situación. Su aplicación a los estudios de predicción de la rugosidad superficial en el caso de maximizar al mismo tiempo la sensibilidad y la especificidad del clasificador inducido de la red Bayesiana, y no solo maximizar la tasa de clasificación correcta. Los intervalos de estos dos objetivos provienen de un m´etodo de estimación honesta de ambos objetivos, como e.g. validación cruzada en k rodajas o bootstrap.---ABSTRACT---The main objective of this PhD Thesis is to go more deeply into the analysis and design of an intelligent system for surface roughness prediction and control in the end-milling machining process, based fundamentally on Bayesian network classifiers, with the aim of developing a methodology that makes easier the design of this type of systems. The system, whose purpose is to make possible the surface roughness prediction and control, consists of a model learnt from experimental data with the aid of Bayesian networks, that will help to understand the dynamic processes involved in the machining and the interactions among the relevant variables. Since artificial neural networks are models widely used in material cutting proceses, we include also an end-milling model using them, where the geometry and hardness of the piecework are introduced as novel variables not studied so far within this context. Thus, an important contribution in this thesis is these two models for surface roughness prediction, that are then compared with respecto to different aspects: influence of the new variables, performance evaluation metrics, interpretability. One of the main problems with Bayesian classifier-based modelling is the understanding of the enormous posterior probabilitiy tables produced. We introduce an explanation method that generates a set of rules obtained from decision trees. Such trees are induced from a simulated data set generated from the posterior probabilities of the class variable, calculated with the Bayesian network learned from a training data set. Finally, we contribute in the multi-objective field in the case that some of the objectives cannot be quantified as real numbers but as interval-valued functions. This often occurs in machine learning applications, especially those based on supervised classification. Specifically, the dominance and Pareto front ideas are extended to this setting. Its application to the surface roughness prediction studies the case of maximizing simultaneously the sensitivity and specificity of the induced Bayesian network classifier, rather than only maximizing the correct classification rate. Intervals in these two objectives come from a honest estimation method of both objectives, like e.g. k-fold cross-validation or bootstrap.

Efficient storage and domain-specific information discovery on semistructured documents

The increasing amount of available semistructured data demands efficient mechanisms to store, process, and search an enormous corpus of data to encourage its global adoption. Current techniques to store semistructured documents either map them to relational databases, or use a combination of flat files and indexes. These two approaches result in a mismatch between the tree-structure of semistructured data and the access characteristics of the underlying storage devices. Furthermore, the inefficiency of XML parsing methods has slowed down the large-scale adoption of XML into actual system implementations. The recent development of lazy parsing techniques is a major step towards improving this situation, but lazy parsers still have significant drawbacks that undermine the massive adoption of XML. ^ Once the processing (storage and parsing) issues for semistructured data have been addressed, another key challenge to leverage semistructured data is to perform effective information discovery on such data. Previous works have addressed this problem in a generic (i.e. domain independent) way, but this process can be improved if knowledge about the specific domain is taken into consideration. ^ This dissertation had two general goals: The first goal was to devise novel techniques to efficiently store and process semistructured documents. This goal had two specific aims: We proposed a method for storing semistructured documents that maps the physical characteristics of the documents to the geometrical layout of hard drives. We developed a Double-Lazy Parser for semistructured documents which introduces lazy behavior in both the pre-parsing and progressive parsing phases of the standard Document Object Model’s parsing mechanism. ^ The second goal was to construct a user-friendly and efficient engine for performing Information Discovery over domain-specific semistructured documents. This goal also had two aims: We presented a framework that exploits the domain-specific knowledge to improve the quality of the information discovery process by incorporating domain ontologies. We also proposed meaningful evaluation metrics to compare the results of search systems over semistructured documents. ^

Efficient Storage and Domain-Specific Information Discovery on Semistructured Documents

The increasing amount of available semistructured data demands efficient mechanisms to store, process, and search an enormous corpus of data to encourage its global adoption. Current techniques to store semistructured documents either map them to relational databases, or use a combination of flat files and indexes. These two approaches result in a mismatch between the tree-structure of semistructured data and the access characteristics of the underlying storage devices. Furthermore, the inefficiency of XML parsing methods has slowed down the large-scale adoption of XML into actual system implementations. The recent development of lazy parsing techniques is a major step towards improving this situation, but lazy parsers still have significant drawbacks that undermine the massive adoption of XML. Once the processing (storage and parsing) issues for semistructured data have been addressed, another key challenge to leverage semistructured data is to perform effective information discovery on such data. Previous works have addressed this problem in a generic (i.e. domain independent) way, but this process can be improved if knowledge about the specific domain is taken into consideration. This dissertation had two general goals: The first goal was to devise novel techniques to efficiently store and process semistructured documents. This goal had two specific aims: We proposed a method for storing semistructured documents that maps the physical characteristics of the documents to the geometrical layout of hard drives. We developed a Double-Lazy Parser for semistructured documents which introduces lazy behavior in both the pre-parsing and progressive parsing phases of the standard Document Object Model's parsing mechanism. The second goal was to construct a user-friendly and efficient engine for performing Information Discovery over domain-specific semistructured documents. This goal also had two aims: We presented a framework that exploits the domain-specific knowledge to improve the quality of the information discovery process by incorporating domain ontologies. We also proposed meaningful evaluation metrics to compare the results of search systems over semistructured documents.

Almetrics: measuring the influence of the media on the social impact of research

Social media is changing the way we interact, present ideas and information and judge the quality of content and contributions. In recent years there have been hundreds of platforms to freely share all kinds of information and connect across networks. These new tools generate activity statistics and interactions among users such as mentions, retweets, conversations, comments on blogs or Facebook; managers references showing popularity ratings of more references shared by other researchers or repositories that generate statistics of visits or downloads of articles. This paper analyzes that have meaning and implications altmetrics, what are its advantages and critical platforms (Almetric.com, ImpactStory, Plos altmetrics, PlumX), reports progress and benefits for authors, publishers and librarians. It concluded that the value of alternative metrics as a complementary tool citation analysis is evident, although it is suggested that you should dig deeper into this issue to unravel the meaning and the potential value of these indicators to assess their potential.

Using multifractals to evaluate oceanographic model skill

We are in an era of unprecedented data volumes generated from observations and model simulations. This is particularly true from satellite Earth Observations (EO) and global scale oceanographic models. This presents us with an opportunity to evaluate large scale oceanographic model outputs using EO data. Previous work on model skill evaluation has led to a plethora of metrics. The paper defines two new model skill evaluation metrics. The metrics are based on the theory of universal multifractals and their purpose is to measure the structural similarity between the model predictions and the EO data. The two metrics have the following advantages over the standard techniques: a) they are scale-free, b) they carry important part of information about how model represents different oceanographic drivers. Those two metrics are then used in the paper to evaluate the performance of the FVCOM model in the shelf seas around the south-west coast of the UK.

Using multifractals to evaluate oceanographic model skill

We are in an era of unprecedented data volumes generated from observations and model simulations. This is particularly true from satellite Earth Observations (EO) and global scale oceanographic models. This presents us with an opportunity to evaluate large scale oceanographic model outputs using EO data. Previous work on model skill evaluation has led to a plethora of metrics. The paper defines two new model skill evaluation metrics. The metrics are based on the theory of universal multifractals and their purpose is to measure the structural similarity between the model predictions and the EO data. The two metrics have the following advantages over the standard techniques: a) they are scale-free, b) they carry important part of information about how model represents different oceanographic drivers. Those two metrics are then used in the paper to evaluate the performance of the FVCOM model in the shelf seas around the south-west coast of the UK.

Evaluating Document Retrieval Methods for Resource Selection in Clustered P2P IR

Resource Selection (or Query Routing) is an important step in P2P IR. Though analogous to document retrieval in the sense of choosing a relevant subset of resources, resource selection methods have evolved independently from those for document retrieval. Among the reasons for such divergence is that document retrieval targets scenarios where underlying resources are semantically homogeneous, whereas peers would manage diverse content. We observe that semantic heterogeneity is mitigated in the clustered 2-tier P2P IR architecture resource selection layer by way of usage of clustering, and posit that this necessitates a re-look at the applicability of document retrieval methods for resource selection within such a framework. This paper empirically benchmarks document retrieval models against the state-of-the-art resource selection models for the problem of resource selection in the clustered P2P IR architecture, using classical IR evaluation metrics. Our benchmarking study illustrates that document retrieval models significantly outperform other methods for the task of resource selection in the clustered P2P IR architecture. This indicates that clustered P2P IR framework can exploit advancements in document retrieval methods to deliver corresponding improvements in resource selection, indicating potential convergence of these fields for the clustered P2P IR architecture.

Designing for User-facing Uncertainty in Everyday Sensing and Prediction

Thesis (Ph.D.)--University of Washington, 2016-07

Visual metrics for the evaluation of sensor data quality in outdoor perception

This paper proposes an experimental study of quality metrics that can be applied to visual and infrared images acquired from cameras onboard an unmanned ground vehicle (UGV). The relevance of existing metrics in this context is discussed and a novel metric is introduced. Selected metrics are evaluated on data collected by a UGV in clear and challenging environmental conditions, represented in this paper by the presence of airborne dust or smoke. An example of application is given with monocular SLAM estimating the pose of the UGV while smoke is present in the environment. It is shown that the proposed novel quality metric can be used to anticipate situations where the quality of the pose estimate will be significantly degraded due to the input image data. This leads to decisions of advantageously switching between data sources (e.g. using infrared images instead of visual images).

Visual metrics for the evaluation of sensor data quality in outdoor perception

This paper proposes an experimental study of quality metrics that can be applied to visual and infrared images acquired from cameras onboard an unmanned ground vehicle (UGV). The relevance of existing metrics in this context is discussed and a novel metric is introduced. Selected metrics are evaluated on data collected by a UGV in clear and challenging environmental conditions, represented in this paper by the presence of airborne dust or smoke.

Metrics to support the evaluation of association rule clustering

Many topics related to association mining have received attention in the research community, especially the ones focused on the discovery of interesting knowledge. A promising approach, related to this topic, is the application of clustering in the pre-processing step to aid the user to find the relevant associative patterns of the domain. In this paper, we propose nine metrics to support the evaluation of this kind of approach. The metrics are important since they provide criteria to: (a) analyze the methodologies, (b) identify their positive and negative aspects, (c) carry out comparisons among them and, therefore, (d) help the users to select the most suitable solution for their problems. Some experiments were done in order to present how the metrics can be used and their usefulness. © 2013 Springer-Verlag GmbH.

Objective metrics for functional evaluation of upper limb during the ADL of drinking: application in SCI

Three-dimensional kinematic analysis provides quantitative assessment of upper limb motion and is used as an outcome measure to evaluate movement disorders. The aim of the present study is to present a set of kinematic metrics for quantifying characteristics of movement performance and the functional status of the subject during the execution of the activity of daily living (ADL) of drinking from a glass. Then, the objective is to apply these metrics in healthy people and a population with cervical spinal cord injury (SCI), and to analyze the metrics ability to discriminate between healthy and pathologic people. 19 people participated in the study: 7 subjects with metameric level C6 tetraplegia, 4 subjects with metameric level C7 tetraplegia and 8 healthy subjects. The movement was recorded with a photogrammetry system. The ADL of drinking was divided into a series of clearly identifiable phases to facilitate analysis. Metrics describing the time of the reaching phase, the range of motion of the joints analyzed, and characteristics of movement performance such as the efficiency, accuracy and smoothness of the distal segment and inter-joint coordination were obtained. The performance of the drinking task was more variable in people with SCI compared to the control group in relation to the metrics measured. Reaching time was longer in SCI groups. The proposed metrics showed capability to discriminate between healthy and pathologic people. Relative deficits in efficiency were larger in SCI people than in controls. These metrics can provide useful information in a clinical setting about the quality of the movement performed by healthy and SCI people during functional activities.

New Framework and Metrics for Economic Evaluation of Medical Technology in the Presence of Heterogeneity and Its Application to the Evaluation of Bariatric Surgery

Thesis (Ph.D.)--University of Washington, 2016-07