Publications

Updated 12 Mar 2015, see also my Google Scholars and Research ID pages.

Authors: Type: Publications:

2015

  • Amalia Z. Berna, X. Rosalind Wang, Julie Cassells, Barry Croft, and Stephen Trowell. Identification of volatile chemicals that can be used to diagnose ratoon stunting disease. , 37:150-157, 2015.
    [Bibtex]
    @article{BernaPASSCT2015,
    Author = {Berna, Amalia Z. and Wang, X. Rosalind and Cassells, Julie and Croft, Barry and Trowell, Stephen},
    Booktitle = {Proceedings of the Australian Society of Sugar Cane Technologists},
    Date-Added = {2016-03-12 03:38:41 +0000},
    Date-Modified = {2016-03-12 03:45:53 +0000},
    Keywords = {VOC, Classification, feature selection},
    Pages = {150-157},
    Title = {Identification of volatile chemicals that can be used to diagnose ratoon stunting disease},
    Volume = {37},
    Year = {2015}}
  • X. Rosalind Wang, Joseph T. Lizier, and Amalia Z. Berna. Feature ſelection for classification using data ſets with ſmall ſample ſizes. In Workshop on learning from ſmall ſample ſizes, conference on knowledge discovery and data mining, Sydney, Australia, aug 2015.
    [Bibtex] [Abstract]
    @inproceedings{WangLSSS2015,
    Abstract = {We are interested in selecting the best features for classification for data sets with a limited number of samples. This problem is prevalent in the biological field where due to various reasons, only small number of samples for each class is collected in experiments. We employ an information-theoretic method to feature selection that is specific to each problem. To estimate the probability density function for data sets with small sample sizes, we use the Kraskov-Grassberger technique that is robust against small data sets and outliers in them. We show in our results that our method can identify the features that produce very good classification performance.},
    Address = {Sydney, Australia},
    Author = {Wang, X. Rosalind and Lizier, Joseph T. and Berna, Amalia Z.},
    Booktitle = {Workshop on Learning from Small Sample Sizes, Conference on Knowledge Discovery and Data Mining},
    Date-Added = {2016-03-12 03:28:26 +0000},
    Date-Modified = {2016-03-12 04:32:02 +0000},
    Keywords = {feature selection, Machine learning, Classification},
    Month = {aug},
    Title = {Feature Selection for Classification using Data Sets with Small Sample Sizes},
    Year = {2015}}

    We are interested in selecting the best features for classification for data sets with a limited number of samples. This problem is prevalent in the biological field where due to various reasons, only small number of samples for each class is collected in experiments. We employ an information-theoretic method to feature selection that is specific to each problem. To estimate the probability density function for data sets with small sample sizes, we use the Kraskov-Grassberger technique that is robust against small data sets and outliers in them. We show in our results that our method can identify the features that produce very good classification performance.

  • Amalia Z. Berna, James McCarthy, X. Rosalind Wang, Kevin Saliba, Florence Bravo, Julie Cassells, Benjamin Padovan, and Stephen Trowell. Biomarkers of malaria infection detected in human breath. In Breath analysis ſummit, Vienna, 2015. IOP Sicence.
    [Bibtex]
    @inproceedings{BernaBAS2015,
    Address = {Vienna},
    Author = {Berna, Amalia Z. and McCarthy, James and Wang, X. Rosalind and Saliba, Kevin and Bravo, Florence and Cassells, Julie and Padovan, Benjamin and Trowell, Stephen},
    Booktitle = {Breath Analysis Summit},
    Date-Modified = {2016-03-12 03:27:22 +0000},
    Keywords = {VOC, Classification},
    Month = sep,
    Publisher = {IOP Sicence},
    Title = {Biomarkers of malaria infection detected in human breath},
    Year = {2015}}
  • [DOI] Amalia Z. Berna, James S. McCarthy, X. Rosalind Wang, Kevin J. Saliba, Florence G. Bravo, Julie Cassells, Benjamin Padovan, and Stephen C. Trowell. Analysis of breath ſpecimens for biomarkers of plasmodium falciparum ınfection. Journal of ınfectious diseases, 212(7):1120-1128, 2015.
    [Bibtex] [Abstract]
    @article{BernaJID2015,
    Abstract = {Currently, the majority of diagnoses of malaria rely on a combination of the patient's clinical presentation and the visualization of parasites on a stained blood film. Breath offers an attractive alternative to blood as the basis for simple, noninvasive diagnosis of infectious diseases. In this study, breath samples were collected from individuals during controlled malaria to determine whether specific malaria-associated volatiles could be detected in breath. We identified 9 compounds whose concentrations varied significantly over the course of malaria: carbon dioxide, isoprene, acetone, benzene, cyclohexanone, and 4 thioethers. The latter group, consisting of allyl methyl sulfide, 1-methylthio-propane, (Z)-1-methylthio-1-propene, and (E)-1-methylthio-1-propene, had not previously been associated with any disease or condition. Before the availability of antimalarial drug treatment, there was evidence of concurrent 48-hour cyclical changes in the levels of both thioethers and parasitemia. When thioether concentrations were subjected to a phase shift of 24 hours, a direct correlation between the parasitemia and volatile levels was revealed. Volatile levels declined monotonically approximately 6.5 hours after initial drug treatment, correlating with clearance of parasitemia. No thioethers were detected in in vitro cultures of Plasmodium falciparum. The metabolic origin of the thioethers is not known, but results suggest that interplay between host and parasite metabolic pathways is involved in the production of these thioethers.},
    Author = {Berna, Amalia Z. and McCarthy, James S. and Wang, X. Rosalind and Saliba, Kevin J. and Bravo, Florence G. and Cassells, Julie and Padovan, Benjamin and Trowell, Stephen C.},
    Date-Modified = {2016-03-12 04:32:59 +0000},
    Doi = {10.1093/infdis/jiv176},
    Eprint = {http://jid.oxfordjournals.org/content/212/7/1120.full.pdf+html},
    Journal = {Journal of Infectious Diseases},
    Keywords = {VOC, Classification},
    Number = {7},
    Pages = {1120-1128},
    Title = {Analysis of Breath Specimens for Biomarkers of Plasmodium falciparum Infection},
    Url = {http://jid.oxfordjournals.org/content/212/7/1120.abstract},
    Volume = {212},
    Year = {2015},
    Bdsk-Url-1 = {http://jid.oxfordjournals.org/content/212/7/1120.abstract},
    Bdsk-Url-2 = {http://dx.doi.org/10.1093/infdis/jiv176}}

    Currently, the majority of diagnoses of malaria rely on a combination of the patient's clinical presentation and the visualization of parasites on a stained blood film. Breath offers an attractive alternative to blood as the basis for simple, noninvasive diagnosis of infectious diseases. In this study, breath samples were collected from individuals during controlled malaria to determine whether specific malaria-associated volatiles could be detected in breath. We identified 9 compounds whose concentrations varied significantly over the course of malaria: carbon dioxide, isoprene, acetone, benzene, cyclohexanone, and 4 thioethers. The latter group, consisting of allyl methyl sulfide, 1-methylthio-propane, (Z)-1-methylthio-1-propene, and (E)-1-methylthio-1-propene, had not previously been associated with any disease or condition. Before the availability of antimalarial drug treatment, there was evidence of concurrent 48-hour cyclical changes in the levels of both thioethers and parasitemia. When thioether concentrations were subjected to a phase shift of 24 hours, a direct correlation between the parasitemia and volatile levels was revealed. Volatile levels declined monotonically approximately 6.5 hours after initial drug treatment, correlating with clearance of parasitemia. No thioethers were detected in in vitro cultures of Plasmodium falciparum. The metabolic origin of the thioethers is not known, but results suggest that interplay between host and parasite metabolic pathways is involved in the production of these thioethers.

  • [DOI] Mikhail Prokopenko, Lionel Barnett, Michael Harré, Joseph T. Lizier, Oliver Obst, and X. Rosalind Wang. Fisher transfer entropy: quantifying the gain in transient sensitivity. Proceedings of the royal ſociety of london a: mathematical, physical and engineering ſciences, 471(2184), 2015.
    [Bibtex] [Abstract]
    @article{ProkopenkoPRSLA2015,
    Abstract = {We introduce a novel measure, Fisher transfer entropy (FTE), which quantifies a gain in sensitivity to a control parameter of a state transition, in the context of another observable source. The new measure captures both transient and contextual qualities of transfer entropy and the sensitivity characteristics of Fisher information. FTE is exemplified for a ferromagnetic two-dimensional lattice Ising model with Glauber dynamics and is shown to diverge at the critical point.},
    Author = {Prokopenko, Mikhail and Barnett, Lionel and Harr{\'e}, Michael and Lizier, Joseph T. and Obst, Oliver and Wang, X. Rosalind},
    Date-Added = {2016-03-12 03:13:23 +0000},
    Date-Modified = {2016-03-12 03:14:13 +0000},
    Doi = {10.1098/rspa.2015.0610},
    Eprint = {http://rspa.royalsocietypublishing.org/content/471/2184/20150610.full.pdf},
    Issn = {1364-5021},
    Journal = {Proceedings of the Royal Society of London A: Mathematical, Physical and Engineering Sciences},
    Keywords = {Fisher information, transfer entropy, Information theory},
    Number = {2184},
    Publisher = {The Royal Society},
    Title = {Fisher transfer entropy: quantifying the gain in transient sensitivity},
    Url = {http://rspa.royalsocietypublishing.org/content/471/2184/20150610},
    Volume = {471},
    Year = {2015},
    Bdsk-Url-1 = {http://rspa.royalsocietypublishing.org/content/471/2184/20150610},
    Bdsk-Url-2 = {http://dx.doi.org/10.1098/rspa.2015.0610}}

    We introduce a novel measure, Fisher transfer entropy (FTE), which quantifies a gain in sensitivity to a control parameter of a state transition, in the context of another observable source. The new measure captures both transient and contextual qualities of transfer entropy and the sensitivity characteristics of Fisher information. FTE is exemplified for a ferromagnetic two-dimensional lattice Ising model with Glauber dynamics and is shown to diverge at the critical point.

  • [PDF] Amalia Z. Berna, X. Rosalind Wang, Julie Cassells, Barry Croft, and Stephen C. Trowell. Diagnostic volatiles to predict ratoon stunting disease. In Australian ſociety of ſugar cane technologists (ASSCT) conference, April 2015.
    [Bibtex] [Abstract]
    @inproceedings{BernaASSCT2015,
    Abstract = {Ratoon Stunting Disease (RSD) caused by Leifsonia xyli subsp. xyli is probably the most significant disease to affect sugarcane. Control relies on strict hygiene, including screening of seed cane, sterilisation of equipment and selective use of partially resistant germplasm. The current standard RSD diagnostic assay is based on an evaporative binding enzyme linked immunosorbent assay (EB-ELISA), which requires sap to be extracted and dried onto an immunoassay plate. Although high throughput is achievable with this method, it is a slow and complicated process. Some species of pathogenic bacteria can be characterized by the volatile chemicals they produce and Leifsonia xyli subsp. xyli is not an exception to this fact. In this paper, we show that at least 10 volatiles increased significantly in levels when the plant was infected with RSD. Gas chromatography-mass spectrometry analysis of the headspace of sugar sap identified some specific alcohols, ketones and aldehyde as signatures of RSD infection. We aim to use this information to training electronic nose sensors to sniff contamination with more sensitivity and accuracy than current ELISA methods.},
    Author = {Amalia Z. Berna and Wang, X. Rosalind and Julie Cassells and Barry Croft and Stephen C. Trowell},
    Booktitle = {Australian Society of Sugar Cane Technologists ({ASSCT}) conference},
    Date-Added = {2014-11-20 00:28:15 +0000},
    Date-Modified = {2016-03-12 04:31:29 +0000},
    Keywords = {VOC, feature selection},
    Month = {April},
    Title = {Diagnostic volatiles to predict ratoon stunting disease},
    Url = {https://www.assct.com.au/},
    Year = {2015}}

    Ratoon Stunting Disease (RSD) caused by Leifsonia xyli subsp. xyli is probably the most significant disease to affect sugarcane. Control relies on strict hygiene, including screening of seed cane, sterilisation of equipment and selective use of partially resistant germplasm. The current standard RSD diagnostic assay is based on an evaporative binding enzyme linked immunosorbent assay (EB-ELISA), which requires sap to be extracted and dried onto an immunoassay plate. Although high throughput is achievable with this method, it is a slow and complicated process. Some species of pathogenic bacteria can be characterized by the volatile chemicals they produce and Leifsonia xyli subsp. xyli is not an exception to this fact. In this paper, we show that at least 10 volatiles increased significantly in levels when the plant was infected with RSD. Gas chromatography-mass spectrometry analysis of the headspace of sugar sap identified some specific alcohols, ketones and aldehyde as signatures of RSD infection. We aim to use this information to training electronic nose sensors to sniff contamination with more sensitivity and accuracy than current ELISA methods.

  • [PDF] [DOI] X. Rosalind Wang, Joseph T. Lizier, Amalia Z. Berna, Florence Bravo, and Stephen C. Trowell. Human breath-print identification by e-nose, using information-theoretic feature selection prior to classification. Sensors and actuators b: chemical, 2015. in press
    [Bibtex] [Abstract]
    @article{WangSNB2015,
    Abstract = {The composition of bodily fluids reflects many aspects of health status of a patient. Breath is anothersample that may be useful for diagnosis of infectious and other diseases. Analysis of breath has theadvantage of being less invasive than analysis of other fluids such as blood and bronchial biopsy. Tworecent studies, using either mass spectrometry or electronic nose (E-nose) technologies, showed there aredefinite ``breath-prints'' that characterised individuals despite temporal variation in internal metabolismand environment. In this study we demonstrate that by employing an information-theoretic featureselection method that is specific to the problem together with machine learning techniques, we candramatically improve (cross-validated) identification of individuals through their breath using a verysmall selected subset of E-nose measurement features. Indeed, we demonstrate here that we can identifythe 10 individuals in this study with perfect accuracy using fewer than 10 features.},
    Author = {X. Rosalind Wang and Joseph T. Lizier and Amalia Z. Berna and Florence Bravo and Stephen C. Trowell},
    Date-Added = {2014-10-02 02:05:29 +0000},
    Date-Modified = {2014-11-24 09:53:14 +0000},
    Doi = {http://dx.doi.org/10.1016/j.snb.2014.09.115},
    Journal = {Sensors and Actuators B: Chemical},
    Keywords = {Classification,Electronic nose,Feature selection,Machine learning,Metabolic phenotype,Mutual information},
    Note = {in press},
    Title = {Human breath-print identification by E-nose, using information-theoretic feature selection prior to classification},
    Url = {http://www.sciencedirect.com/science/article/pii/S0925400514012052},
    Year = {2015},
    Bdsk-Url-1 = {http://www.sciencedirect.com/science/article/pii/S0925400514012052},
    Bdsk-Url-2 = {http://dx.doi.org/10.1016/j.snb.2014.09.115}}

    The composition of bodily fluids reflects many aspects of health status of a patient. Breath is anothersample that may be useful for diagnosis of infectious and other diseases. Analysis of breath has theadvantage of being less invasive than analysis of other fluids such as blood and bronchial biopsy. Tworecent studies, using either mass spectrometry or electronic nose (E-nose) technologies, showed there aredefinite ``breath-prints'' that characterised individuals despite temporal variation in internal metabolismand environment. In this study we demonstrate that by employing an information-theoretic featureselection method that is specific to the problem together with machine learning techniques, we candramatically improve (cross-validated) identification of individuals through their breath using a verysmall selected subset of E-nose measurement features. Indeed, we demonstrate here that we can identifythe 10 individuals in this study with perfect accuracy using fewer than 10 features.

2014

  • Jennifer M. Miller, X. Rosalind Wang, Joseph T. Lizier, Mikhail Prokopenko, and Louis F. Rossi. Measuring ınformation dynamics in ſwarms. In Mikhail Prokopenko, editor, Guided ſelf-organization: ınception, pages 343-364. Springer, 2014.
    [Bibtex] [Abstract]
    @incollection{miller14,
    Abstract = {We propose a novel, information theoretic characterization of dynamics within swarms, through explicitly measuring the extent of collective communications and tracing collectivememory. These elements of distributed computation provide complementary views into the capacity for swarm coherence and reorganization. The approach deals with both global and local information dynamics ultimately discovering diverse ways in which an individual's location within the group is related to its information processing role.},
    Author = {Miller, Jennifer M. and Wang, X. Rosalind and Lizier, Joseph T. and Prokopenko, Mikhail and Rossi, Louis F.},
    Booktitle = {Guided Self-Organization: Inception},
    Date-Added = {2014-09-05 06:53:23 +0000},
    Date-Modified = {2014-11-24 09:56:07 +0000},
    Editor = {Prokopenko, Mikhail},
    Keywords = {Complex systems,Computational Biology,Information theory,transfer entropy,Social and behavioral sciences,},
    Pages = {343-364},
    Publisher = {Springer},
    Title = {Measuring Information Dynamics in Swarms},
    Year = {2014}}

    We propose a novel, information theoretic characterization of dynamics within swarms, through explicitly measuring the extent of collective communications and tracing collectivememory. These elements of distributed computation provide complementary views into the capacity for swarm coherence and reorganization. The approach deals with both global and local information dynamics ultimately discovering diverse ways in which an individual's location within the group is related to its information processing role.

  • Yu Sun, Louis F. Rossi, Chien-Chung Shen, Jennifer Miller, X. Rosalind Wang, Joseph T Lizier, Mikhail Prokopenko, and Upul Senanayake. Information transfer in ſwarms with leaders. In Collective ıntelligence 2014, June 2014.
    [Bibtex]
    @inproceedings{Sun2014,
    Author = {Yu Sun and Louis F. Rossi and Chien-Chung Shen and Jennifer Miller and X. Rosalind Wang and Joseph T Lizier and Mikhail Prokopenko and Upul Senanayake},
    Booktitle = {Collective Intelligence 2014},
    Date-Added = {2014-09-05 06:50:05 +0000},
    Date-Modified = {2014-11-24 09:49:59 +0000},
    Keywords = {Information theory,Complex systems,Computational Biology,Social and behavioral sciences},
    Month = {June},
    Title = {Information Transfer in Swarms with Leaders},
    Year = {2014}}
  • Thomas Nowotny, Amalia Z. Berna, Russell Binions, X. Rosalind Wang, Joseph T. Lizier, Mikhail Prokopenko, and Stephen Trowell. Feature selection in enose applications. In Proceedings of the 1st ınternational workshop on odor ſpaces, September 2014.
    [Bibtex]
    @inproceedings{Nowotny2014,
    Author = {Thomas Nowotny and Amalia Z. Berna and Russell Binions and X. Rosalind Wang and Joseph T. Lizier and Mikhail Prokopenko and Stephen Trowell},
    Booktitle = {Proceedings of the 1st International Workshop on Odor Spaces},
    Date-Added = {2014-09-05 06:47:17 +0000},
    Date-Modified = {2014-11-24 09:04:38 +0000},
    Keywords = {feature selection,},
    Month = {September},
    Title = {Feature selection in Enose applications},
    Year = {2014}}
  • X. Rosalind Wang, Amalia Z. Berna, Joseph T. Lizier, Florence Bravo, and Stephen C. Trowell. Can a machine learn to identify individuals through breath?. In 15th ınternational meeting on chemical ſensors (IMCS), Buenos Aires, Argentina, March 2014.
    [Bibtex]
    @inproceedings{WangIMCS2014,
    Address = {Buenos Aires, Argentina},
    Author = {X. Rosalind Wang and Amalia Z. Berna and Joseph T. Lizier and Florence Bravo and Stephen C. Trowell},
    Booktitle = {15th International Meeting on Chemical Sensors ({IMCS})},
    Date-Added = {2014-09-05 06:39:11 +0000},
    Date-Modified = {2014-11-24 09:53:52 +0000},
    Keywords = {Electronic nose,Machine learning,Classification,feature selection},
    Month = {March},
    Title = {Can a machine learn to identify individuals through breath?},
    Year = {2014}}
  • [DOI] X. Rosalind Wang, Joseph T. Lizier, Thomas Nowotny, Amalia Z. Berna, Mikhail Prokopenko, and Stephen C. Trowell. Feature ſelection for chemical ſensor arrays using mutual ınformation. PLoS ONE, 9(3):e89840, March 2014.
    [Bibtex] [Abstract]
    @article{WangPLOS2014,
    Abstract = {We address the problem of feature selection for classifying a diverse set of chemicals using an array of metal oxide sensors. Our aim is to evaluate a filter approach to feature selection with reference to previous work, which used a wrapper approach on the same data set, and established best features and upper bounds on classification performance. We selected feature sets that exhibit the maximal mutual information with the identity of the chemicals. The selected features closely match those found to perform well in the previous study using a wrapper approach to conduct an exhaustive search of all permitted feature combinations. By comparing the classification performance of support vector machines (using features selected by mutual information) with the performance observed in the previous study, we found that while our approach does not always give the maximum possible classification performance, it always selects features that achieve classification performance approaching the optimum obtained by exhaustive search. We performed further classification using the selected feature set with some common classifiers and found that, for the selected features, Bayesian Networks gave the best performance. Finally, we compared the observed classification performances with the performance of classifiers using randomly selected features. We found that the selected features consistently outperformed randomly selected features for all tested classifiers. The mutual information filter approach is therefore a computationally efficient method for selecting near optimal features for chemical sensor arrays.},
    Author = {X. Rosalind Wang and Joseph T. Lizier and Thomas Nowotny and Amalia Z. Berna and Mikhail Prokopenko and Stephen C. Trowell},
    Date-Modified = {2014-11-24 09:54:09 +0000},
    Doi = {10.1371/journal.pone.0089840},
    Journal = {{PLoS ONE}},
    Keywords = {feature selection,Mutual information,Classification,Machine learning},
    Month = {March},
    Number = {3},
    Pages = {e89840},
    Publisher = {Public Library of Science},
    Title = {Feature Selection for Chemical Sensor Arrays Using Mutual Information},
    Volume = {9},
    Year = {2014},
    Bdsk-Url-1 = {http://dx.doi.org/10.1371/journal.pone.0089840}}

    We address the problem of feature selection for classifying a diverse set of chemicals using an array of metal oxide sensors. Our aim is to evaluate a filter approach to feature selection with reference to previous work, which used a wrapper approach on the same data set, and established best features and upper bounds on classification performance. We selected feature sets that exhibit the maximal mutual information with the identity of the chemicals. The selected features closely match those found to perform well in the previous study using a wrapper approach to conduct an exhaustive search of all permitted feature combinations. By comparing the classification performance of support vector machines (using features selected by mutual information) with the performance observed in the previous study, we found that while our approach does not always give the maximum possible classification performance, it always selects features that achieve classification performance approaching the optimum obtained by exhaustive search. We performed further classification using the selected feature set with some common classifiers and found that, for the selected features, Bayesian Networks gave the best performance. Finally, we compared the observed classification performances with the performance of classifiers using randomly selected features. We found that the selected features consistently outperformed randomly selected features for all tested classifiers. The mutual information filter approach is therefore a computationally efficient method for selecting near optimal features for chemical sensor arrays.

2013

  • Oliver M. Cliff, Joseph T. Lizier, X. Rosalind Wang, Peter Wang, Oliver Obst, and Mikhail Prokopenko. Towards quantifying ınteraction networks in a football match. In Proceedings of the RoboCup 2013 ſymposium, 2013.
    Best paper award for the best theoretical contribution.
    [Bibtex] [Abstract]
    @inproceedings{CliffRobocup2013,
    Abstract = {We present several novel methods quantifying dynamic interactions in simulated football games. These interactions are captured in directed networks that represent significant coupled dynamics, detected information-theoretically. The model-free approach measures information dynamics of both pair-wise players' interactions as well as local tactical contests produced during RoboCup 2D Simulation League games. This analysis involves computation of information transfer and storage, relating the information transfer to responsiveness of the players and the team, and the information storage within the team to the team's rigidity and lack of tactical flexibility. The resultant directed networks (interaction diagrams) and the measures of responsiveness and rigidity reveal implicit interactions, across teams, that may be delayed and/or long-ranged. The analysis was verified with a number of experiments, identifying the zones of the most intense competition and the extent of interactions.},
    Author = {Oliver M. Cliff and Joseph T. Lizier and X. Rosalind Wang and Peter Wang and Oliver Obst and Mikhail Prokopenko},
    Booktitle = {Proceedings of the {RoboCup 2013} Symposium},
    Date-Modified = {2014-11-25 07:26:05 +0000},
    Keywords = {Information theory,transfer entropy},
    Title = {Towards Quantifying Interaction Networks in a Football Match},
    Wwwnote = {<strong>Best paper award for the best theoretical contribution.</strong>},
    Year = {2013}}

    We present several novel methods quantifying dynamic interactions in simulated football games. These interactions are captured in directed networks that represent significant coupled dynamics, detected information-theoretically. The model-free approach measures information dynamics of both pair-wise players' interactions as well as local tactical contests produced during RoboCup 2D Simulation League games. This analysis involves computation of information transfer and storage, relating the information transfer to responsiveness of the players and the team, and the information storage within the team to the team's rigidity and lack of tactical flexibility. The resultant directed networks (interaction diagrams) and the measures of responsiveness and rigidity reveal implicit interactions, across teams, that may be delayed and/or long-ranged. The analysis was verified with a number of experiments, identifying the zones of the most intense competition and the extent of interactions.

2012

  • [DOI] X. Rosalind Wang, Jennifer M. Miller, Joseph T. Lizier, Mikhail Prokopenko, and Louis F. Rossi. Quantifying and tracing information cascades in swarms.. PLoS ONE, 7(7):e40084, 2012.
    [Bibtex] [Abstract]
    @article{WangPLOS2012,
    Abstract = {We propose a novel, information-theoretic, characterisation of cascades within the spatiotemporal dynamics of swarms, explicitly measuring the extent of collective communications. This is complemented by dynamic tracing of collective memory, as another element of distributed computation, which represents capacity for swarm coherence. The approach deals with both global and local information dynamics, ultimately discovering diverse ways in which an individual's spatial position is related to its information processing role. It also allows us to contrast cascades that propagate conflicting information with waves of coordinated motion. Most importantly, our simulation experiments provide the first direct information-theoretic evidence (verified in a simulation setting) for the long-held conjecture that the information cascades occur in waves rippling through the swarm. Our experiments also exemplify how features of swarm dynamics, such as cascades' wavefronts, can be filtered and predicted. We observed that maximal information transfer tends to follow the stage with maximal collective memory, and principles like this may be generalised in wider biological and social contexts.},
    Author = {Wang, X. Rosalind and Miller, Jennifer M. and Lizier, Joseph T. and Prokopenko, Mikhail and Rossi, Louis F.},
    Date-Modified = {2014-11-25 00:15:43 +0000},
    Doi = {10.1371/journal.pone.0089840},
    Editor = {de Polavieja, Gonzalo G.},
    File = {::},
    Issn = {1932-6203},
    Journal = {{PLoS ONE}},
    Keywords = {Complex systems,Computational Biology,Information theory,Social and behavioral sciences,transfer entropy},
    Month = jan,
    Number = {7},
    Pages = {e40084},
    Publisher = {Public Library of Science},
    Title = {{Quantifying and tracing information cascades in swarms.}},
    Url = {http://dx.plos.org/10.1371/journal.pone.0040084},
    Volume = {7},
    Year = {2012},
    Bdsk-Url-1 = {http://dx.plos.org/10.1371/journal.pone.0040084},
    Bdsk-Url-2 = {http://dx.doi.org/10.1371/journal.pone.0089840}}

    We propose a novel, information-theoretic, characterisation of cascades within the spatiotemporal dynamics of swarms, explicitly measuring the extent of collective communications. This is complemented by dynamic tracing of collective memory, as another element of distributed computation, which represents capacity for swarm coherence. The approach deals with both global and local information dynamics, ultimately discovering diverse ways in which an individual's spatial position is related to its information processing role. It also allows us to contrast cascades that propagate conflicting information with waves of coordinated motion. Most importantly, our simulation experiments provide the first direct information-theoretic evidence (verified in a simulation setting) for the long-held conjecture that the information cascades occur in waves rippling through the swarm. Our experiments also exemplify how features of swarm dynamics, such as cascades' wavefronts, can be filtered and predicted. We observed that maximal information transfer tends to follow the stage with maximal collective memory, and principles like this may be generalised in wider biological and social contexts.

2011

  • X. Rosalind Wang, Joseph T Lizier, and Josh Wall. Information transfer analysis for ongoing commissioning of buildings. In Autonomous agents and multiagent ſystems, Taipei, Taiwan, 2011.
    [Bibtex]
    @inproceedings{WangAAMS2011,
    Address = {Taipei, Taiwan},
    Author = {Wang, X. Rosalind and Lizier, Joseph T and Wall, Josh},
    Booktitle = {Autonomous Agents and Multiagent Systems},
    Date-Added = {2014-09-06 03:36:22 +0000},
    Date-Modified = {2014-11-24 09:57:32 +0000},
    Keywords = {transfer entropy,anomaly},
    Title = {Information Transfer Analysis for Ongoing Commissioning of Buildings},
    Year = {2011}}
  • Samuel R West, Ying Guo, X. Rosalind Wang, and Joshua Wall. Automated fault detection and diagnosis of hvac subsystems using statistical machine learning. In V Soebarto, H Bennetts, P Bannister, PC Thomas, and D Leach, editors, 12th ınternational conference of the ınternational building performance ſimulation association, 2011.
    Award for best research paper
    [Bibtex]
    @inproceedings{west2011automated,
    Author = {West, Samuel R and Guo, Ying and Wang, X. Rosalind and Wall, Joshua},
    Booktitle = {12th International Conference of the International Building Performance Simulation Association},
    Date-Added = {2014-09-06 03:34:32 +0000},
    Date-Modified = {2014-11-25 07:33:41 +0000},
    Editor = {Soebarto, V and Bennetts, H and Bannister, P and Thomas, PC and Leach, D},
    Keywords = {anomaly,transfer entropy,Machine learning,},
    Title = {Automated fault detection and diagnosis of HVAC subsystems using statistical machine learning},
    Wwwnote = {<strong>Award for best research paper</strong>},
    Year = {2011}}
  • X. Rosalind Wang, Jennifer M. Miller, Joseph T. Lizier, Mikhail Prokopenko, and Louis F. Rossi. Measuring ınformation ſtorage and transfer in ſwarms. In European conference on artificial life (ECAL), 2011.
    [Bibtex] [Abstract]
    @inproceedings{WangECAL2011,
    Abstract = {Spatial aggregation of animal groups give individuals many benefits that they would not be able to obtain otherwise. One of the key questions in the study of these animal groups, or ``swarms'', concerns the way in which information is propagated through the group. In this paper, we examine this propagation using an information-theoretic framework of distributed computation. Swarm dynamics is interpreted as a type of distributed computation. Two localized informationtheoretic measures (active information storage and transfer entropy) are adapted to the task of tracing the information dynamics in a kinematic context. The observed types of swarm dynamics, as well as transitions among these types, are shown to coincide with well-marked local and global optima of the proposed measures. Specifically, active information storage tends to maximize as the swarm is becoming less fragmented and the kinematic history begins to strongly inform an observer about the next state. The peak of transfer entropy is observed to appear at the final stages of merging of swarm fragments, near the ``edge of chaos'' where the system actively computes its next stable configuration. Both measures tend to minimize for either unstable or static swarm configurations. The results here show these measures can be applied to non-trivial models, most importantly, they can tell us about the dynamics within these model where we can not rely on visual intuitions.},
    Author = {Wang, X. Rosalind and Miller, Jennifer M. and Lizier, Joseph T. and Prokopenko, Mikhail and Rossi, Louis F.},
    Booktitle = {European Conference on Artificial Life ({ECAL})},
    Date-Added = {2014-09-06 03:32:12 +0000},
    Date-Modified = {2014-11-24 09:58:06 +0000},
    Keywords = {transfer entropy,Complex systems,Computational Biology,Social and behavioral sciences,},
    Title = {Measuring Information Storage and Transfer in Swarms},
    Year = {2011}}

    Spatial aggregation of animal groups give individuals many benefits that they would not be able to obtain otherwise. One of the key questions in the study of these animal groups, or ``swarms'', concerns the way in which information is propagated through the group. In this paper, we examine this propagation using an information-theoretic framework of distributed computation. Swarm dynamics is interpreted as a type of distributed computation. Two localized informationtheoretic measures (active information storage and transfer entropy) are adapted to the task of tracing the information dynamics in a kinematic context. The observed types of swarm dynamics, as well as transitions among these types, are shown to coincide with well-marked local and global optima of the proposed measures. Specifically, active information storage tends to maximize as the swarm is becoming less fragmented and the kinematic history begins to strongly inform an observer about the next state. The peak of transfer entropy is observed to appear at the final stages of merging of swarm fragments, near the ``edge of chaos'' where the system actively computes its next stable configuration. Both measures tend to minimize for either unstable or static swarm configurations. The results here show these measures can be applied to non-trivial models, most importantly, they can tell us about the dynamics within these model where we can not rely on visual intuitions.

  • X. Rosalind Wang, Joseph T. Lizier, and Mikhail Prokopenko. Fisher information at the edge of chaos in random boolean networks. Artificial life, 17(4):315-329, 2011.
    [Bibtex] [Abstract]
    @article{WangALife2011,
    Abstract = {We study the order-chaos phase transition in random Boolean networks (RBNs), which have been used as models of gene regulatory networks. In particular we seek to characterize the phase diagram in information-theoretic terms, focusing on the effect of the control parameters (activity level and connectivity). Fisher information, which measures how much system dynamics can reveal about the control parameters, offers a natural interpretation of the phase diagram in RBNs. We report that this measure is maximized near the order-chaos phase transitions in RBNs, since this is the region where the system is most sensitive to its parameters. Furthermore, we use this study of RBNs to clarify the relationship between Shannon and Fisher information measures.},
    Author = {X. Rosalind Wang and Joseph T. Lizier and Mikhail Prokopenko},
    Date-Added = {2014-09-06 03:26:12 +0000},
    Date-Modified = {2014-11-24 10:00:21 +0000},
    Journal = {Artificial Life},
    Keywords = {Fisher information,random Boolean network,Information theory,order parameter, phase transitions},
    Number = {4},
    Pages = {315-329},
    Title = {Fisher information at the edge of chaos in random Boolean networks},
    Volume = {17},
    Year = {2011}}

    We study the order-chaos phase transition in random Boolean networks (RBNs), which have been used as models of gene regulatory networks. In particular we seek to characterize the phase diagram in information-theoretic terms, focusing on the effect of the control parameters (activity level and connectivity). Fisher information, which measures how much system dynamics can reveal about the control parameters, offers a natural interpretation of the phase diagram in RBNs. We report that this measure is maximized near the order-chaos phase transitions in RBNs, since this is the region where the system is most sensitive to its parameters. Furthermore, we use this study of RBNs to clarify the relationship between Shannon and Fisher information measures.

  • [DOI] Mikhail Prokopenko, Joseph T. Lizier, Oliver Obst, and X. Rosalind Wang. Relating Fisher information to order parameters. Physical review e, 84(4):41116, 2011.
    [Bibtex] [Abstract]
    @article{ProkopenkoPRE2011,
    Abstract = {We study phase transitions and relevant order parameters via statistical estimation theory using the Fisher information matrix. The assumptions that we make limit our analysis to order parameters representable as a negative derivative of thermodynamic potential over some thermodynamic vari- able. Nevertheless, the resulting representation is sufficiently general and explicitly relates elements of the Fisher information matrix to the rate of change in the corresponding order parameters. The obtained relationships allow us to identify, in particular,
    second-order phase transitions via diver- gences of individual elements of the Fisher information matrix. A computational study of random Boolean networks (RBNs) supports the derived relationships, illustrating that Fisher information of the magnetization bias (that is,
    activity level) is peaked in finite-size networks at the critical points, and the maxima increase with the network size. The framework presented here reveals the basic thermodynamic reasons behind similar empirical observations reported previously. The study highlights the generality of Fisher information as a measure that can be applied to a broad range of systems, particularly those where the determination of order parameters is cumbersome.},
    Author = {Mikhail Prokopenko and Joseph T. Lizier and Oliver Obst and X. Rosalind Wang},
    Bib2Html_Sel = {selected},
    Date-Added = {2011-11-11 03:04:54 +0000},
    Date-Modified = {2014-12-07 00:08:09 +0000},
    Doi = {10.1103/PhysRevE.84.041116},
    Journal = {Physical Review E},
    Keywords = {phase transitions, Fisher information, order parameter, thermodynamic potential, free entropy, random Boolean network,},
    Number = {4},
    Pages = {041116},
    Title = {Relating {Fisher} information to order parameters},
    Volume = {84},
    Year = {2011},
    Bdsk-Url-1 = {http://dx.doi.org/10.1103/PhysRevE.84.041116}}

    We study phase transitions and relevant order parameters via statistical estimation theory using the Fisher information matrix. The assumptions that we make limit our analysis to order parameters representable as a negative derivative of thermodynamic potential over some thermodynamic vari- able. Nevertheless, the resulting representation is sufficiently general and explicitly relates elements of the Fisher information matrix to the rate of change in the corresponding order parameters. The obtained relationships allow us to identify, in particular, second-order phase transitions via diver- gences of individual elements of the Fisher information matrix. A computational study of random Boolean networks (RBNs) supports the derived relationships, illustrating that Fisher information of the magnetization bias (that is, activity level) is peaked in finite-size networks at the critical points, and the maxima increase with the network size. The framework presented here reveals the basic thermodynamic reasons behind similar empirical observations reported previously. The study highlights the generality of Fisher information as a measure that can be applied to a broad range of systems, particularly those where the determination of order parameters is cumbersome.

  • [DOI] Raja Jurdak, X. Rosalind Wang, Oliver Obst, and Philip Valencia. Wireless ſensor network anomalies: diagnosis and detection ſtrategies, volume 10, chapter 12, pages 309-325. Springer, Berlin, Heidelberg, 2011.
    [Bibtex] [Abstract]
    @inbook{JurdakIBSE2011,
    Abstract = {Wireless Sensor Networks (WSNs) can experience problems (anomalies) during deployment, due to dynamic environmental factors or node hardware and software failures. These anomalies demand reliable detection strategies for supporting long term and/or large scale WSN deployments. Several strategies have been proposed for detecting specific subsets of WSN anomalies, yet there is still a need for more comprehensive anomaly detection strategies that jointly address network, node, and data level anomalies. This chapter examines WSN anomalies from an intelligent-based system perspective, covering anomalies that arise at the network, node and data levels. It generalizes a simple process for diagnosing anomalies in WSNs for detection, localization, and root cause determination. A survey of existing anomaly detection strategies also reveals their major design choices,
    including architecture and user support, and yields guidelines for tailoring new anomaly detection strategies to specific WSN application requirements.},
    Address = {Berlin, Heidelberg},
    Affiliation = {CSIRO ICT Centre, Australia},
    Author = {Jurdak, Raja and Wang, X. Rosalind and Obst, Oliver and Valencia, Philip},
    Bib2Html_Sel = {selected},
    Booktitle = {Intelligence-Based Systems Engineering},
    Chapter = {12},
    Date-Added = {2011-03-26 16:05:50 +1100},
    Date-Modified = {2014-11-24 09:55:27 +0000},
    Doi = {10.1007/978-3-642-17931-0_12},
    Editor = {Kacprzyk, Janusz and Jain, Lakhmi C. and Tolk, Andreas and Jain, Lakhmi C.},
    Isbn = {978-3-642-17931-0},
    Keyword = {wireless sensor networks, anomaly detection, diagnosis},
    Keywords = {sensor networks,Machine learning,},
    Pages = {309--325},
    Publisher = {Springer},
    Series = {Intelligent Systems Reference Library},
    Title = {Wireless Sensor Network Anomalies: Diagnosis and Detection Strategies},
    Volume = {10},
    Year = {2011},
    Bdsk-Url-1 = {http://dx.doi.org/10.1007/978-3-642-17931-0_12}}

    Wireless Sensor Networks (WSNs) can experience problems (anomalies) during deployment, due to dynamic environmental factors or node hardware and software failures. These anomalies demand reliable detection strategies for supporting long term and/or large scale WSN deployments. Several strategies have been proposed for detecting specific subsets of WSN anomalies, yet there is still a need for more comprehensive anomaly detection strategies that jointly address network, node, and data level anomalies. This chapter examines WSN anomalies from an intelligent-based system perspective, covering anomalies that arise at the network, node and data levels. It generalizes a simple process for diagnosing anomalies in WSNs for detection, localization, and root cause determination. A survey of existing anomaly detection strategies also reveals their major design choices, including architecture and user support, and yields guidelines for tailoring new anomaly detection strategies to specific WSN application requirements.

2010

  • X. Rosalind Wang, Joseph T. Lizier, and Mikhail Prokopenko. A Fisher ınformation ſtudy of phase transitions in Random Boolean Networks. In Alife 2010, 2010.
    [Bibtex] [Abstract]
    @inproceedings{WangALife2010,
    Abstract = {We study the order-chaos phase transition in random Boolean networks (RBNs), which have been used as models of gene regulatory networks. In particular we seek to characterise the phase diagram in information-theoretic terms, focussing on the effect of the control parameters (activity level and connectivity). Fisher information, which measures how much system dynamics reveal about its parameters, offers a natural interpretation of the phase diagram in RBNs. We report that this measure is maximised near the critical state in the order-chaos phase transitions in RBNs, since this is the region where the system is most sensitive to its parameters. Furthermore, we use this study of RBNs to clarify the relationship between Shannon and Fisher information measures.},
    Author = {Wang, X. Rosalind and Lizier, Joseph T. and Prokopenko, Mikhail},
    Booktitle = {ALife 2010},
    Date-Modified = {2014-11-24 10:00:37 +0000},
    File = {::},
    Keywords = {Fisher information,Information theory,random Boolean network,phase transitions},
    Title = {A {Fisher} Information Study of Phase Transitions in {Random Boolean Networks}},
    Year = {2010}}

    We study the order-chaos phase transition in random Boolean networks (RBNs), which have been used as models of gene regulatory networks. In particular we seek to characterise the phase diagram in information-theoretic terms, focussing on the effect of the control parameters (activity level and connectivity). Fisher information, which measures how much system dynamics reveal about its parameters, offers a natural interpretation of the phase diagram in RBNs. We report that this measure is maximised near the critical state in the order-chaos phase transitions in RBNs, since this is the region where the system is most sensitive to its parameters. Furthermore, we use this study of RBNs to clarify the relationship between Shannon and Fisher information measures.

2009

  • X. Rosalind Wang, George Mathews, Don Price, and Mikhail Prokopenko. Optimising ſensor layouts for direct measurement of discrete variables. In Self-adaptive and ſelf-organizing ſystems, (SASO'09). third IEEE ınternational conference on, pages 92-102, 2009.
    [Bibtex]
    @inproceedings{WangSASO2009,
    Author = {Wang, X. Rosalind and Mathews, George and Price, Don and Prokopenko, Mikhail},
    Booktitle = {Self-Adaptive and Self-Organizing Systems, ({SASO}'09). Third {IEEE} International Conference on},
    Date-Modified = {2014-11-24 02:55:58 +0000},
    Pages = {92--102},
    Title = {Optimising Sensor Layouts for Direct Measurement of Discrete Variables},
    Year = {2009}}

2008

  • [PDF] X. Rosalind Wang, Vadim Gerasimov, Mikhail Prokopenko, and Astrid Zeman. Predicting ſeasonal mixed-mode time ſeries. not published, 2008.
    [Bibtex] [Abstract]
    @misc{WangPred2008,
    Abstract = {Underlying phase space dynamics of many time series can be characterized by mixed-mode periodic and chaotic orbits. Mixed-mode dynamics create a serious pattern recognition challenge for the analysis and forecasting of such time series. We propose here a novel method, which uses a combination of attractor reconstruction and Bayesian Network modeling. Dimensionality analysis is used to find embedding dimensions of the data and possible dimensions of underlying attractors. The correct dimensions allow us to build non-linear Bayesian models, used in predicting chaotic time-series. The developed algorithm is successfully applied to Australian electricity demand data.},
    Author = {X. Rosalind Wang and Vadim Gerasimov and Mikhail Prokopenko and Astrid Zeman},
    Date-Added = {2014-11-22 05:05:07 +0000},
    Date-Modified = {2014-11-22 23:27:04 +0000},
    Howpublished = {not published},
    Title = {Predicting Seasonal Mixed-mode Time Series},
    Year = {2008}}

    Underlying phase space dynamics of many time series can be characterized by mixed-mode periodic and chaotic orbits. Mixed-mode dynamics create a serious pattern recognition challenge for the analysis and forecasting of such time series. We propose here a novel method, which uses a combination of attractor reconstruction and Bayesian Network modeling. Dimensionality analysis is used to find embedding dimensions of the data and possible dimensions of underlying attractors. The correct dimensions allow us to build non-linear Bayesian models, used in predicting chaotic time-series. The developed algorithm is successfully applied to Australian electricity demand data.

  • Oliver Obst, X. Rosalind Wang, and Mikhail Prokopenko. Using echo ſtate networks for anomaly detection in underground coal mines. In Proceedings of the ınternational conference on ınformation processing in ſensor networks (IPSN 2008), pages 219-229. IEEE Computer Society, 2008.
    [Bibtex] [Abstract]
    @inproceedings{ObstIPSN2008,
    Abstract = {We investigate the problem of identifying anomalies in monitoring critical gas concentrations using a sensor network in an underground coal mine. In this domain, one of the main problems is a provision of mine specific anomaly detection, with cyclical (moving) instead of flatline (static) alarm threshold levels. An additional practical difficulty in modelling a specific mine is the lack of fully labelled data of normal and abnormal situations. We present an approach addressing these difficulties based on echo state networks learning mine specific anomalies when only normal data is available. Echo state networks utilize incremental updates driven by new sensor readings, thus enabling a detection of anomalies at any time during the sensor network operation. We evaluate this approach against a benchmark -- Bayes Network based anomaly detection, and observe that the quality of the overall predictions is comparable to the benchmark. However, the echo state networks maintain the same level of predictive accuracy for data from multiple sources. Therefore, the ability of echo state networks to model dynamical systems make this approach more suitable for anomaly detection and predictions in sensor networks. },
    Author = {Oliver Obst and X. Rosalind Wang and Mikhail Prokopenko},
    Bib2Html_Sel = {selected},
    Booktitle = {Proceedings of the International Conference on Information Processing in Sensor Networks ({IPSN} 2008)},
    Date-Added = {2008-01-30 12:18:05 +1100},
    Date-Modified = {2014-11-24 09:01:02 +0000},
    Isbn = {978-0-7695-3157-1},
    Keywords = {echo state networks, recurrent neural networks, sensor networks, bayesian networks, anomaly},
    Month = apr,
    Pages = {219--229},
    Publisher = {IEEE Computer Society},
    Read = {Yes},
    Title = {Using Echo State Networks for Anomaly Detection in Underground Coal Mines},
    Year = {2008}}

    We investigate the problem of identifying anomalies in monitoring critical gas concentrations using a sensor network in an underground coal mine. In this domain, one of the main problems is a provision of mine specific anomaly detection, with cyclical (moving) instead of flatline (static) alarm threshold levels. An additional practical difficulty in modelling a specific mine is the lack of fully labelled data of normal and abnormal situations. We present an approach addressing these difficulties based on echo state networks learning mine specific anomalies when only normal data is available. Echo state networks utilize incremental updates driven by new sensor readings, thus enabling a detection of anomalies at any time during the sensor network operation. We evaluate this approach against a benchmark -- Bayes Network based anomaly detection, and observe that the quality of the overall predictions is comparable to the benchmark. However, the echo state networks maintain the same level of predictive accuracy for data from multiple sources. Therefore, the ability of echo state networks to model dynamical systems make this approach more suitable for anomaly detection and predictions in sensor networks.

  • [DOI] X. Rosalind Wang, Joseph T. Lizier, Oliver Obst, Mikhail Prokopenko, and Peter Wang. Spatiotemporal anomaly detection in gas monitoring ſensor networks. In Roberto Verdone, editor, Wireless ſensor networks, volume 4913, pages 90-105. Springer, Berlin, Heidelberg, 2008.
    [Bibtex] [Abstract]
    @incollection{WangEWSN2008,
    Abstract = {In this paper, we use Bayesian Networks as a means for unsupervised learning and anomaly (event) detection in gas monitoring sensor networks for underground coal mines. We show that the Bayesian Network model can learn cyclical baselines for gas concentrations, thus reducing false alarms usually caused by flatline thresholds. Further, we show that the system can learn dependencies between changes of concentration in different gases and at multiple locations. We define and identify new types of events that can occur in a sensor network. In particular, we analyse joint events in a group of sensors based on learning the Bayesian model of the system, contrasting these events with merely aggregating single events. We demonstrate that anomalous events in individual gas data might be explained if considered jointly with the changes in other gases. Vice versa, a network-wide spatiotemporal anomaly may be detected even if individual sensor readings were within their thresholds. The presented Bayesian approach to spatiotemporal anomaly detection is applicable to a wide range of sensor networks. },
    Address = {Berlin, Heidelberg},
    Author = {X. Rosalind Wang and Joseph T. Lizier and Oliver Obst and Mikhail Prokopenko and Peter Wang},
    Bib2Html_Sel = {selected},
    Booktitle = {Wireless Sensor Networks},
    Date-Added = {2008-01-31 17:30:05 +1100},
    Date-Modified = {2014-11-24 09:03:59 +0000},
    Doi = {10.1007/978-3-540-77690-1},
    Editor = {Roberto Verdone},
    Keywords = {anomaly,bayesian networks,sensor networks},
    Month = feb,
    Pages = {90--105},
    Publisher = {Springer},
    Series = {Lecture Notes in Computer Science},
    Title = {Spatiotemporal Anomaly Detection in Gas Monitoring Sensor Networks},
    Url = {http://www.springerlink.com/content/7253h68412q7l083/},
    Volume = {4913},
    Year = {2008},
    Bdsk-Url-1 = {http://www.springerlink.com/content/7253h68412q7l083/},
    Bdsk-Url-2 = {http://dx.doi.org/10.1007/978-3-540-77690-1}}

    In this paper, we use Bayesian Networks as a means for unsupervised learning and anomaly (event) detection in gas monitoring sensor networks for underground coal mines. We show that the Bayesian Network model can learn cyclical baselines for gas concentrations, thus reducing false alarms usually caused by flatline thresholds. Further, we show that the system can learn dependencies between changes of concentration in different gases and at multiple locations. We define and identify new types of events that can occur in a sensor network. In particular, we analyse joint events in a group of sensors based on learning the Bayesian model of the system, contrasting these events with merely aggregating single events. We demonstrate that anomalous events in individual gas data might be explained if considered jointly with the changes in other gases. Vice versa, a network-wide spatiotemporal anomaly may be detected even if individual sensor readings were within their thresholds. The presented Bayesian approach to spatiotemporal anomaly detection is applicable to a wide range of sensor networks.

2006

  • X. Rosalind Wang, Suresh Kumar, Fabio Ramos, and Tobias Kaupp. Probabilistic classification of hyperspectral images by learning nonlinear dimensionality reduction mapping. In Information fusion, 2006 9th ınternational conference on, pages 1-8. IEEE,, 2006.
    [Bibtex]
    @inproceedings{wang2006probabilistic,
    Author = {Wang, X. Rosalind and Kumar, Suresh and Ramos, Fabio and Kaupp, Tobias},
    Booktitle = {Information Fusion, 2006 9th International Conference on},
    Date-Added = {2014-09-06 03:38:28 +0000},
    Date-Modified = {2014-11-23 05:30:11 +0000},
    Organization = {IEEE},
    Pages = {1--8},
    Title = {Probabilistic classification of hyperspectral images by learning nonlinear dimensionality reduction mapping},
    Year = {2006}}

2005

  • X. Rosalind Wang and Fabio T. Ramos. Applying structural em in autonomous planetary exploration missions using hyperspectral image spectroscopy. In Robotics and automation, 2005. ıcra 2005. proceedings of the 2005 ıeee ınternational conference on, pages 4284-4289. IEEE,, 2005.
    [Bibtex]
    @inproceedings{wang2005sem,
    Author = {Wang, X. Rosalind and Ramos, Fabio T.},
    Booktitle = {Robotics and Automation, 2005. ICRA 2005. Proceedings of the 2005 IEEE International Conference on},
    Date-Added = {2014-09-06 03:39:26 +0000},
    Date-Modified = {2014-11-23 05:30:21 +0000},
    Organization = {IEEE},
    Pages = {4284--4289},
    Title = {Applying structural EM in autonomous planetary exploration missions using hyperspectral image spectroscopy},
    Year = {2005}}
  • X. Rosalind Wang, Adrian J. Brown, and Ben Upcroft. Applying incremental em to bayesian classifiers in the learning of hyperspectral remote sensing data. In Information fusion, 2005 8th ınternational conference on, volume 1, page 8--pp. IEEE,, 2005.
    [Bibtex]
    @inproceedings{wang2005iem,
    Author = {Wang, X. Rosalind and Brown, Adrian J. and Upcroft, Ben},
    Booktitle = {Information Fusion, 2005 8th International Conference on},
    Date-Added = {2014-09-06 03:38:52 +0000},
    Date-Modified = {2014-11-23 05:38:12 +0000},
    Organization = {IEEE},
    Pages = {8--pp},
    Title = {Applying incremental EM to bayesian classifiers in the learning of hyperspectral remote sensing data},
    Volume = {1},
    Year = {2005}}

Page generated using papercite plugin for WordPress originally written by Benjamin Piwowarski, with changes by Oliver Obst.