We present a flexible deep convolutional neural network method for the analysis of arbitrary size... more We present a flexible deep convolutional neural network method for the analysis of arbitrary sized graph structures representing molecules. This method, which makes use of the Lipinski RDKit module, an open-source cheminformatics software, enables the incorporation of any global molecular (such as molecular charge and molecular weight) and local (such as atom hybridization and bond orders) information. In this paper, we show that this method significantly outperforms another recently proposed method based on deep convolutional neural networks on several datasets that are studied. Several best practices for training deep convolutional neural networks on chemical datasets are also highlighted within the article, such as how to select the information to be included in the model, how to prevent overfitting and how unbalanced classes in the data can be handled.
Studies in health technology and informatics, 2000
MedView is a joint project with participants from oral medicine and computer science. The aim of ... more MedView is a joint project with participants from oral medicine and computer science. The aim of the project is to build a large database from patient examinations and produce computerized tools to extend, view, and analyze the contents of the database. The contents of the data base is based on a formalization of health-care processes and clinical knowledge in oral medicine harmonized within the network SOMNET. We give an overview of the current status of the MedView project and discuss background and future directions.
Visual Analytics for Homeland Defense and Security, 2009
ABSTRACT Monitoring the surveillance of large sea areas normally involves the analysis of huge qu... more ABSTRACT Monitoring the surveillance of large sea areas normally involves the analysis of huge quantities of heterogeneous data from multiple sources (radars, cameras, automatic identification systems, reports, etc.). The rapid identification of anomalous behavior or any ...
2010 Seventh International Conference on Computer Graphics, Imaging and Visualization, 2010
ABSTRACT Many approaches for anomaly detection use statistical based methods that build profiles ... more ABSTRACT Many approaches for anomaly detection use statistical based methods that build profiles of normality. In these cases, anomalies are defined as deviations from normal models build from representative data. Detection systems based solely on these approaches typically generate high false alarm rates due to the difficulty of creating flawless models. In order to support the comprehension, validation and update of such models, this paper is devoted to the visualization of normal behavioral models of sea traffic and their usability evaluation. First, we present geographical projections of the different probability density functions that represent the normal traffic behavior and second, we outline results from a usability assessment carried out in order to evaluate the ability of such visualizations to support representative tasks related to the establishment of normal situational picture.
Studies in Health Technology and Informatics, 2005
How can information technology be used to model and handle clinical knowledge in everyday work so... more How can information technology be used to model and handle clinical knowledge in everyday work so that clinicians can more systematically learn from collected clinical data? Ontology is a crucial element in answering this question. Based on nearly ten years of clinical experience, fundamental requirements of an ontology for oral medicine are presented. The use of the proposed W3C standards RDF and OWL for the design and implementation of an ontology for oral medicine is then described and discussed. The reported work contributes to knowledge representation in oral medicine by presenting the pioneering work towards an ontology for oral medicine using RDF/OWL, thereby testing the latter on a new domain.
How can information technology be used to model and handle clinical knowledge in everyday work so... more How can information technology be used to model and handle clinical knowledge in everyday work so that clinicians can more systematically learn from collected clini-cal data? Ontology is a crucial element in answering this question. Based on nearly ten years of clinical experience, fundamental requirements of an ontology for oral medicine are presented. The use of the proposed W3C standards RDF and OWL for the design and implementation of an ontology for oral medicine is then described and discussed. The reported work contributes to knowledge representation in oral medicine by presenting the pioneering work towards an ontology for oral medicine using RDF/OWL, thereby testing the latter on a new domain.
This paper presents a novel application of the theory of con-formal prediction for distribution-i... more This paper presents a novel application of the theory of con-formal prediction for distribution-independent on-line learn-ing and anomaly detection. We exploit the fact that confor-mal predictors give valid prediction sets at specified confi-dence levels under the relatively weak assumption that the (normal) training data together with (normal) observations to be predicted have been generated from the same distri-bution. If the actual observation is not included in the pos-sibly empty prediction set, it is classified as anomalous at the corresponding significance level. Interpreting the signif-icance level as an upper bound of the probability that a nor-mal observation is mistakenly classified as anomalous, we can conveniently adjust the sensitivity to anomalies while controlling the rate of false alarms without having to find any application specific thresholds. The proposed method has been evaluated in the domain of sea surveillance using recorded data assumed to be normal. The va...
We present a new self-organizing algorithm for classification of a data that combines and extends... more We present a new self-organizing algorithm for classification of a data that combines and extends the strengths of several common machine learning algorithms, such as algorithms in self-organizing neural networks, ensemble methods and deep neural networks. The increased expression power is combined with the explanation power of self-organizing networks. Our algorithm outperforms both deep neural networks and ensembles of deep neural networks. For our evaluation case, we use production monitoring data from a complex steel manufacturing process, where data is both high-dimensional and has many nonlinear interdependencies. In addition to the improved prediction score, the algorithm offers a new deep-learning based approach for how computational resources can be focused in data exploration, since the algorithm points out areas of the input space that are more challenging to learn.
Data Science offers a set of powerful approaches for making new discoveries from large and comple... more Data Science offers a set of powerful approaches for making new discoveries from large and complex data sets. It combines aspects of mathematics, statistics, machine learning, etc. to turn vast amounts of data into new insights and knowledge. However, the sole use of automatic data science techniques for large amounts of complex data limits the human user’s possibilities in the discovery process, since the user is estranged from the process of data exploration. This chapter describes the importance of Information Visualization (InfoVis) and visual analytics (VA) within data science and how interactive visualization can be used to support analysis and decision-making, empowering and complementing data science methods. Moreover, we review perceptual and cognitive aspects, together with design and evaluation methodologies for InfoVis and VA.
In the defense domain, to estimate if a target is threatening and to which degree is a complex ta... more In the defense domain, to estimate if a target is threatening and to which degree is a complex task, that is typically carried out by human operators due to the high risks and uncertainties associated. To their aid, different support systems have been implemented to analyze the data and provide recommendations for actions. Since the ultimate responsibility lies in human operators, it is of utmost importance that they trust and know how to use these systems, as well as have an understanding of their inner workings, strengths and limitations. This paper presents, first, a formative user study to characterize how air traffic operators carry out threat evaluation related tasks. Grounded in these findings and in guidelines found in the literature, we present a transparent and highly interactive prototype that aims at reducing operator's cognitive load and support threat assessment activities. The literature review provided on design guidelines, the outcomes of the user study, the des...
Abstract The competitiveness in the manufacturing industry raises demands for using recent data a... more Abstract The competitiveness in the manufacturing industry raises demands for using recent data analysis algorithms for manufacturing process development. Data-driven analysis enables extraction of novel knowledge from already existing sensors and data, which is necessary for advanced manufacturing process refinement involving aged machinery. Improved data analysis enables factories to stay competitive against newer factories, but without any hefty investment. In large manufacturing operations, the dependencies between data are highly complex and therefore very difficult to analyse manually. This paper applies a deep learning approach, using a recurrent neural network with long short term memory cells together with an attention mechanism to model the dependencies between the measured product shape, as measured before the most critical manufacturing operation, and the final product quality. Our approach predicts the ratio of flawed products already before the critical operation with an AUC-ROC score of 0.85, i.e., we can detect more than 80 % of all flawed products while having less than 25 % false positive predictions (false alarms). In contrast to previous deep learning approaches, our method shows how the recurrent neural network reasons about the input shape, using the attention mechanism to point out which parts of the product shape that have the highest influence on the predictions. Such information is crucial for both process developers, in order to understand and improve the process, and for process operators who can use the information to learn how to better trust the predictions and control the process.
Advances in Intelligent Systems and Computing, 2016
This paper investigates the nature of user participation in the process of designing fighter airc... more This paper investigates the nature of user participation in the process of designing fighter aircraft cockpits. The role of the users, i.e. pilots, in the design of cockpit interfaces is explored. We present the results of an on-line questionnaire with twelve designers of cockpit interfaces for fighter aircraft. The results show that the designers have highlighted the need for more opportunities to observe the pilots, and they wish to obtain more information and ideas from them. Moreover, a larger involvement from users as examiners and testers in the evaluation process was desirable. Access to users was considered unproblematic and the risk of misunderstandings was reported to be low. Moreover, the designers did not support the idea that users should design or take design decisions.
We present a flexible deep convolutional neural network method for the analysis of arbitrary size... more We present a flexible deep convolutional neural network method for the analysis of arbitrary sized graph structures representing molecules. This method, which makes use of the Lipinski RDKit module, an open-source cheminformatics software, enables the incorporation of any global molecular (such as molecular charge and molecular weight) and local (such as atom hybridization and bond orders) information. In this paper, we show that this method significantly outperforms another recently proposed method based on deep convolutional neural networks on several datasets that are studied. Several best practices for training deep convolutional neural networks on chemical datasets are also highlighted within the article, such as how to select the information to be included in the model, how to prevent overfitting and how unbalanced classes in the data can be handled.
Studies in health technology and informatics, 2000
MedView is a joint project with participants from oral medicine and computer science. The aim of ... more MedView is a joint project with participants from oral medicine and computer science. The aim of the project is to build a large database from patient examinations and produce computerized tools to extend, view, and analyze the contents of the database. The contents of the data base is based on a formalization of health-care processes and clinical knowledge in oral medicine harmonized within the network SOMNET. We give an overview of the current status of the MedView project and discuss background and future directions.
Visual Analytics for Homeland Defense and Security, 2009
ABSTRACT Monitoring the surveillance of large sea areas normally involves the analysis of huge qu... more ABSTRACT Monitoring the surveillance of large sea areas normally involves the analysis of huge quantities of heterogeneous data from multiple sources (radars, cameras, automatic identification systems, reports, etc.). The rapid identification of anomalous behavior or any ...
2010 Seventh International Conference on Computer Graphics, Imaging and Visualization, 2010
ABSTRACT Many approaches for anomaly detection use statistical based methods that build profiles ... more ABSTRACT Many approaches for anomaly detection use statistical based methods that build profiles of normality. In these cases, anomalies are defined as deviations from normal models build from representative data. Detection systems based solely on these approaches typically generate high false alarm rates due to the difficulty of creating flawless models. In order to support the comprehension, validation and update of such models, this paper is devoted to the visualization of normal behavioral models of sea traffic and their usability evaluation. First, we present geographical projections of the different probability density functions that represent the normal traffic behavior and second, we outline results from a usability assessment carried out in order to evaluate the ability of such visualizations to support representative tasks related to the establishment of normal situational picture.
Studies in Health Technology and Informatics, 2005
How can information technology be used to model and handle clinical knowledge in everyday work so... more How can information technology be used to model and handle clinical knowledge in everyday work so that clinicians can more systematically learn from collected clinical data? Ontology is a crucial element in answering this question. Based on nearly ten years of clinical experience, fundamental requirements of an ontology for oral medicine are presented. The use of the proposed W3C standards RDF and OWL for the design and implementation of an ontology for oral medicine is then described and discussed. The reported work contributes to knowledge representation in oral medicine by presenting the pioneering work towards an ontology for oral medicine using RDF/OWL, thereby testing the latter on a new domain.
How can information technology be used to model and handle clinical knowledge in everyday work so... more How can information technology be used to model and handle clinical knowledge in everyday work so that clinicians can more systematically learn from collected clini-cal data? Ontology is a crucial element in answering this question. Based on nearly ten years of clinical experience, fundamental requirements of an ontology for oral medicine are presented. The use of the proposed W3C standards RDF and OWL for the design and implementation of an ontology for oral medicine is then described and discussed. The reported work contributes to knowledge representation in oral medicine by presenting the pioneering work towards an ontology for oral medicine using RDF/OWL, thereby testing the latter on a new domain.
This paper presents a novel application of the theory of con-formal prediction for distribution-i... more This paper presents a novel application of the theory of con-formal prediction for distribution-independent on-line learn-ing and anomaly detection. We exploit the fact that confor-mal predictors give valid prediction sets at specified confi-dence levels under the relatively weak assumption that the (normal) training data together with (normal) observations to be predicted have been generated from the same distri-bution. If the actual observation is not included in the pos-sibly empty prediction set, it is classified as anomalous at the corresponding significance level. Interpreting the signif-icance level as an upper bound of the probability that a nor-mal observation is mistakenly classified as anomalous, we can conveniently adjust the sensitivity to anomalies while controlling the rate of false alarms without having to find any application specific thresholds. The proposed method has been evaluated in the domain of sea surveillance using recorded data assumed to be normal. The va...
We present a new self-organizing algorithm for classification of a data that combines and extends... more We present a new self-organizing algorithm for classification of a data that combines and extends the strengths of several common machine learning algorithms, such as algorithms in self-organizing neural networks, ensemble methods and deep neural networks. The increased expression power is combined with the explanation power of self-organizing networks. Our algorithm outperforms both deep neural networks and ensembles of deep neural networks. For our evaluation case, we use production monitoring data from a complex steel manufacturing process, where data is both high-dimensional and has many nonlinear interdependencies. In addition to the improved prediction score, the algorithm offers a new deep-learning based approach for how computational resources can be focused in data exploration, since the algorithm points out areas of the input space that are more challenging to learn.
Data Science offers a set of powerful approaches for making new discoveries from large and comple... more Data Science offers a set of powerful approaches for making new discoveries from large and complex data sets. It combines aspects of mathematics, statistics, machine learning, etc. to turn vast amounts of data into new insights and knowledge. However, the sole use of automatic data science techniques for large amounts of complex data limits the human user’s possibilities in the discovery process, since the user is estranged from the process of data exploration. This chapter describes the importance of Information Visualization (InfoVis) and visual analytics (VA) within data science and how interactive visualization can be used to support analysis and decision-making, empowering and complementing data science methods. Moreover, we review perceptual and cognitive aspects, together with design and evaluation methodologies for InfoVis and VA.
In the defense domain, to estimate if a target is threatening and to which degree is a complex ta... more In the defense domain, to estimate if a target is threatening and to which degree is a complex task, that is typically carried out by human operators due to the high risks and uncertainties associated. To their aid, different support systems have been implemented to analyze the data and provide recommendations for actions. Since the ultimate responsibility lies in human operators, it is of utmost importance that they trust and know how to use these systems, as well as have an understanding of their inner workings, strengths and limitations. This paper presents, first, a formative user study to characterize how air traffic operators carry out threat evaluation related tasks. Grounded in these findings and in guidelines found in the literature, we present a transparent and highly interactive prototype that aims at reducing operator's cognitive load and support threat assessment activities. The literature review provided on design guidelines, the outcomes of the user study, the des...
Abstract The competitiveness in the manufacturing industry raises demands for using recent data a... more Abstract The competitiveness in the manufacturing industry raises demands for using recent data analysis algorithms for manufacturing process development. Data-driven analysis enables extraction of novel knowledge from already existing sensors and data, which is necessary for advanced manufacturing process refinement involving aged machinery. Improved data analysis enables factories to stay competitive against newer factories, but without any hefty investment. In large manufacturing operations, the dependencies between data are highly complex and therefore very difficult to analyse manually. This paper applies a deep learning approach, using a recurrent neural network with long short term memory cells together with an attention mechanism to model the dependencies between the measured product shape, as measured before the most critical manufacturing operation, and the final product quality. Our approach predicts the ratio of flawed products already before the critical operation with an AUC-ROC score of 0.85, i.e., we can detect more than 80 % of all flawed products while having less than 25 % false positive predictions (false alarms). In contrast to previous deep learning approaches, our method shows how the recurrent neural network reasons about the input shape, using the attention mechanism to point out which parts of the product shape that have the highest influence on the predictions. Such information is crucial for both process developers, in order to understand and improve the process, and for process operators who can use the information to learn how to better trust the predictions and control the process.
Advances in Intelligent Systems and Computing, 2016
This paper investigates the nature of user participation in the process of designing fighter airc... more This paper investigates the nature of user participation in the process of designing fighter aircraft cockpits. The role of the users, i.e. pilots, in the design of cockpit interfaces is explored. We present the results of an on-line questionnaire with twelve designers of cockpit interfaces for fighter aircraft. The results show that the designers have highlighted the need for more opportunities to observe the pilots, and they wish to obtain more information and ideas from them. Moreover, a larger involvement from users as examiners and testers in the evaluation process was desirable. Access to users was considered unproblematic and the risk of misunderstandings was reported to be low. Moreover, the designers did not support the idea that users should design or take design decisions.
Uploads