2018

  • E. Chatzaki, M. Panagopoulou, M. Karaglani, I. Balgkouranidou, V. Vasilakakis, E. Biziota, T. Koukaki, E. Karamitrousis, E. Nena, I. Tsamardinos, G. Kolios, E. Lianidou, and S. Kakolyris, «Circulating cell free DNA in Breast cancer: size profiling, levels and methylation patterns lead to prognostic and predictive classifiers,» (to appear) Oncogene , 2018.
    [Summary]

  • M. Adamou, G. Antoniou, E. Greasidou, V. Lagani, P. Charonyktakis, I. Tsamardinos, and M. Doyle, «Toward Automatic Risk Assessment to Support Suicide Prevention,» Crisis, 2018.
    [Summary]

    Background: Suicide has been considered an important public health issue for years and is one of the main causes of death worldwide. Despite prevention strategies being applied, the rate of suicide has not changed substantially over the past decades. Suicide risk has proven extremely difficult to assess for medical specialists, and traditional methodologies deployed have been ineffective. Advances in machine learning make it possible to attempt to predict suicide with the analysis of relevant data aiming to inform clinical practice. Aims: We aimed to (a) test our artificial intelligence based, referral-centric methodology in the context of the National Health Service (NHS), (b) determine whether statistically relevant results can be derived from data related to previous suicides, and (c) develop ideas for various exploitation strategies. Method: The analysis used data of patients who died by suicide in the period 2013–2016 including both structured data and free-text medical notes, necessitating the deployment of state-of-the-art machine learning and text mining methods. Limitations: Sample size is a limiting factor for this study, along with the absence of non-suicide cases. Specific analytical solutions were adopted for addressing both issues. Results and Conclusion: The results of this pilot study indicate that machine learning shows promise for predicting within a specified period which people are most at risk of taking their own life at the time of referral to a mental health service.

  • [PDF] [DOI] M. Tsagris, V. Lagani, and I. Tsamardinos, «12Test Feature selection for high-dimensional temporal data,» BMC Bioinformatics, vol. 19, iss. 1, p. 17, 2018.
    [Summary]

    Feature selection is commonly employed for identifying collectively-predictive biomarkers and biosignatures; it facilitates the construction of small statistical models that are easier to verify, visualize, and comprehend while providing insight to the human expert. In this work, we extend established constrained-based, feature-selection methods to high-dimensional “omics” temporal data, where the number of measurements is orders of magnitude larger than the sample size. The extension required the development of conditional independence tests for temporal and/or static variables conditioned on a set of temporal variables. The algorithm is able to return multiple, equivalent solution subsets of variables, scale to tens of thousands of features, and outperform or be on par with existing methods depending on the analysis task specifics. The use of this algorithm is suggested for variable selection with high-dimensional temporal data.

  • [DOI] K. Lakiotaki, N. Vorniotakis, M. Tsagris, G. Georgakopoulos, and I. Tsamardinos, «BioDataome: a collection of uniformly preprocessed and automatically annotated datasets for data-driven biology,» Database, vol. 2018, iss. bay011, 2018.
    [Summary]

    Biotechnology revolution generates a plethora of omics data with an exponential growth pace. Therefore, biological data mining demands automatic, ‘high quality’ curation efforts to organize biomedical knowledge into online databases. BioDataome is a database of uniformly preprocessed and disease-annotated omics data with the aim to promote and accelerate the reuse of public data. We followed the same preprocessing pipeline for each biological mart (microarray gene expression, RNA-Seq gene expression and DNA methylation) to produce ready for downstream analysis datasets and automatically annotated them with disease-ontology terms. We also designate datasets that share common samples and automatically discover control samples in case-control studies. Currently, BioDataome includes ∼5600 datasets, ∼260 000 samples spanning ∼500 diseases and can be easily used in large-scale massive experiments and meta-analysis. All datasets are publicly available for querying and downloading via BioDataome web application. We demonstrate BioDataome’s utility by presenting exploratory data analysis examples. We have also developed BioDataome R package found in: https://github.com/mensxmachina/BioDataome/. Database URL: http://dataome.mensxmachina.org/

  • M. Adamou, G. Antoniou, E. Greassidou, V. Lagani, P. Charonyktakis, and I. Tsamardinos, «Mining Free-Text Medical Notes for Suicide Risk Assessment,» in SETN, 2018.
    [Summary]

  • M. Adamou, G. Antoniou, E. Greassidou, V. Lagani, P. Charonyktakis, and I. Tsamardinos, «Towards Automatic Risk Assessment to Support Suicide Prevention,» Accepted to Crisis: The Journal of Crisis Intervention and Suicide Prevention, 2018.
    [Summary]

  • [DOI] M. Markaki, I. Tsamardinos, A. Langhammer, V. Lagani, K. Hveem, and O. D. Røe, «A Validated Clinical Risk Prediction Model for Lung Cancer in Smokers of All Ages and Exposure Types: A HUNT Study.,» EBioMedicine, 2018.
    [Summary]

    Lung cancer causes \textgreater1\textperiodcentered6 million deaths annually, with early diagnosis being paramount to effective treatment. Here we present a validated risk assessment model for lung cancer screening. The prospective HUNT2 population study in Norway examined 65,237 people aged \textgreater20years in 1995-97. After a median of 15\textperiodcentered2years, 583 lung cancer cases had been diagnosed; 552 (94\textperiodcentered7\%) ever-smokers and 31 (5\textperiodcentered3\%) never-smokers. We performed multivariable analyses of 36 candidate risk predictors, using multiple imputation of missing data and backwards feature selection with Cox regression. The resulting model was validated in an independent Norwegian prospective dataset of 45,341 ever-smokers, in which 675 lung cancers had been diagnosed after a median follow-up of 11\textperiodcentered6years. Our final HUNT Lung Cancer Model included age, pack-years, smoking intensity, years since smoking cessation, body mass index, daily cough, and hours of daily indoors exposure to smoke. External validation showed a 0\textperiodcentered879 concordance index (95\% CI [0\textperiodcentered866-0\textperiodcentered891]) with an area under the curve of 0\textperiodcentered87 (95\% CI [0\textperiodcentered85-0\textperiodcentered89]) within 6years. Only 22\% of ever-smokers would need screening to identify 81\textperiodcentered85\% of all lung cancers within 6years. Our model of seven variables is simple, accurate, and useful for screening selection.

  • [DOI] M. Tsagris, G. Borboudakis, V. Lagani, and I. Tsamardinos, «Constraint-based causal discovery with mixed data,» International Journal of Data Science and Analytics, 2018.
    [Summary]

    We address the problem of constraint-based causal discovery with mixed data types, such as (but not limited to) continuous, binary, multinomial and or-dinal variables. We use likelihood-ratio tests based on appropriate regression models, and show how to derive symmetric conditional independence tests. Such tests can then be directly used by existing constraint-based methods with mixed data, such as the PC and FCI algorithms for learning Bayesian networks and maximal ancestral graphs respectively. In experiments on simu-lated Bayesian networks, we employ the PC algorithm with different conditional independence tests for mixed data, and show that the proposed approach outperforms alternatives in terms of learning accuracy.

2017

  • [DOI] K. Siomos, E. Papadaki, I. Tsamardinos, K. Kerkentzes, M. Koygioylis, and C. Trakatelli, «Prothrombotic and Endothelial Inflammatory Markers in Greek Patients with Type 2 Diabetes Compared to Non-Diabetics,» Endocrinology & Metabolic Syndrome, vol. 6, iss. 1, 2017.
    [Summary]

    Objective: To evaluate specific factors of coagulation and endothelial inflammatory markers namely, thrombomodulin, soluble receptor of the protein C (sEPCR), factor VIII, plasminogen activator inhibitor 1, Von Willebrandt factor, fibrinogen, fibrinogen dimers (d-dimers), high sensitivity C-reactive protein and homocysteine in a subset of Greek subjects with and without Type 2 (T2) Diabetes. Design: 84 subjects, of which 44 patients with T2 diabetes, were included in the randomized comparative prospective cross sectional study. The subjects were split into a Τ2 diabetics group and a group of healthy controls of similar age, anthropometric profiles and similar gender distribution. Results: A total of 47 variables and biomarkers together with indicators for metabolic profiles, clinical history, as well as detailed anthropometric profiles and traditional risk factors, were evaluated. Dipeptidyl peptidase-4 (DPP4), Insulin, use of Sulfonylurea, high HBA1c and glucose levels, were clearly statistically differentiated in the two groups, while no other biomarkers including the new potential indicators were found to be different. High values of thrombomodulin and homocysteine were correlated with a rise in creatinine and thus seem to affect renal function in the diabetic patients group while in the non-diabetics group the correlations are different with sEPCR having a relative strong negative correlation in renal function as measured with The Modification of Diet in Renal Disease, in agreement with the latest international findings. Conclusions: The presence of T2 diabetes in conjunction with age clearly correlates with problems in renal function, thrombomodulin and homocysteine could serve as indicators for renal damage in diabetics but not in healthy individuals. sEPCR on the other hand could be a potential generic indicator for renal damage. Thrombomodulin and sEPCR as prothombotic agents, did not show any indication that they can be utilised as markers for the prevention and/or treatment of thrombotic complications in diabetic patients.

  • [DOI] G. Papoutsoglou, G. Athineou, V. Lagani, I. Xanthopoulos, A. Schmidt, S. éliás, J. Tegnér, and I. Tsamardinos, «SCENERY: a web application for (causal) network reconstruction from cytometry data,» Nucleic Acids Research, vol. 37, p. D412–D416, 2017.
    [Summary]

  • G. Orfanoudaki, M. Markaki, K. Chatzi, I. Tsamardinos, and A. Economou, «MatureP: prediction of secreted proteins with exclusive information from their mature regions,» Scientific Reports, vol. 7, iss. 1, p. 3263, 2017.
    [Summary]

  • K. Tsirlis, V. Lagani, S. Triantafillou, and I. Tsamardinos, «On Scoring Maximal Ancestral Graphs with the Max-Min Hill Climbing Algorithm,» in 23d ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD 2017), 2017.
    [Summary]

    We consider the problem of causal struc-ture learning in presence of latent confounders. We pro-pose a hybrid method, MAG Max-Min Hill-Climbing (M 3 HC) that takes as input a data set of continuous variables, assumed to follow a multivariate Gaussian distribution, and outputs the best fitting maximal an-cestral graph. M 3 HC builds upon a previously proposed method, namely GSMAG, by introducing a constraint-based first phase that greatly reduces the space of struc-tures to investigate. We show on simulated data that the proposed algorithm greatly improves on GSMAG, and compares positively against FCI and cFCI, two well known constraint-based approaches for causal-network reconstruction in presence of latent confounders.

  • M. Tsagris, G. Borboudakis, V. Lagani, and I. Tsamardinos, «Constraint-based Causal Discovery with Mixed Data,» in 23d ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD 2017), 2017.
    [Summary]

    We address the problem of constraint-based causal discovery with mixed data types, such as (but not limited to) continuous, binary, multinomial and or-dinal variables. We use likelihood-ratio tests based on appropriate regression models, and show how to derive symmetric conditional independence tests. Such tests can then be directly used by existing constraint-based methods with mixed data, such as the PC and FCI algorithms for learning Bayesian networks and maximal ancestral graphs respectively. In experiments on simu-lated Bayesian networks, we employ the PC algorithm with different conditional independence tests for mixed data, and show that the proposed approach outperforms alternatives in terms of learning accuracy.

  • V. Lagani, G. Athineou, A. Farcomeni, M. Tsagris, and I. Tsamardinos, «Feature Selection with the R Package MXM: Discovering Multiple, Statistically-Equivalent, Predictive Feature Subsets,» Journal of Statistical Software, vol. 80, iss. 7, 2017.
    [Summary]

    The statistically equivalent signature (SES) algorithm is a method for feature selection inspired by the principles of constrained-based learning of Bayesian Networks. Most of the currently available feature-selection methods return only a single subset of features, supposedly the one with the highest predictive power. We argue that in several domains multiple subsets can achieve close to maximal predictive accuracy, and that arbitrarily providing only one has several drawbacks. The SES method attempts to identify multiple, predictive feature subsets whose performances are statistically equivalent. Under that respect SES subsumes and extends previous feature selection algorithms, like the max-min parent children algorithm. SES is implemented in an homonym function included in the R package MXM, standing for mens ex machina, meaning ‘mind from the machine’ in Latin. The MXM implementation of SES handles several data-analysis tasks, namely classification, regression and survival analysis. In this paper we present the SES algorithm, its implementation, and provide examples of use of the SES function in R. Furthermore, we analyze three publicly available data sets to illustrate the equivalence of the signatures retrieved by SES and to contrast SES against the state-of-the-art feature selection method LASSO. Our results provide initial evidence that the two methods perform comparably well in terms of predictive accuracy and that multiple, equally predictive signatures are actually present in real world data.

  • [DOI] S. Triantafillou, V. Lagani, C. Heinze-Deml, A. Schmidt, J. Tegner, and I. Tsamardinos, «Predicting Causal Relationships from Biological Data: Applying Automated Casual Discovery on Mass Cytometry Data of Human Immune Cells,» Triantafillou S, Lagani V, Heinze-Deml C, Schmidt A, Tegner J, Tsamardinos I. Predicting Causal Relationships from Biological Data: Applying Automated Causal Discovery on Mass Cytometry Data of Human Immune Cells. Scientific Reports. 2017;7:12724. doi:10., vol. 7, p. 12724, 2017.
    [Summary]

    Learning the causal relationships that define a molecular system allows us to predict how the system will respond to different interventions. Distinguishing causality from mere association typically requires randomized experiments. Methods for automated causal discovery from limited experiments exist, but have so far rarely been tested in systems biology applications. In this work, we apply state-of-the art causal discovery methods on a large collection of public mass cytometry data sets, measuring intra-cellular signaling proteins of the human immune system and their response to several perturbations. We show how different experimental conditions can be used to facilitate causal discovery, and apply two fundamental methods that produce context-specific causal predictions. Causal predictions were reproducible across independent data sets from two different studies, but often disagree with the KEGG pathway databases. Within this context, we discuss the caveats we need to overcome for automated causal discovery to become a part of the routine data analysis in systems biology.

  • I. Tsamardinos, E. Greasidou, M. Tsagris, and G. Borboudakis, «Bootstrapping the Out-of-sample Predictions for Efficient and Accurate Cross-Validation,» , 2017.
    [Summary]

    Cross-Validation (CV), and out-of-sample performance-estimation protocols in general, are often employed both for (a) selecting the optimal combination of algorithms and values of hyper-parameters (called a configuration) for producing the final predictive model, and (b) estimating the predictive performance of the final model. However, the cross-validated performance of the best configuration is optimistically biased. We present an efficient bootstrap method that corrects for the bias, called Bootstrap Bias Corrected CV (BBC-CV). BBC-CV’s main idea is to bootstrap the whole process of selecting the best-performing configuration on the out-of-sample predictions of each configuration, without additional training of models. In comparison to the alternatives, namely the nested cross-validation and a method by Tibshirani and Tibshirani, BBC-CV is computationally more efficient, has smaller variance and bias, and is applicable to any metric of performance (accuracy, AUC, concordance index, mean squared error). Subsequently, we employ again the idea of bootstrapping the out-of-sample predictions to speed up the CV process. Specifically, using a bootstrap-based hypothesis test we stop training of models on new folds of statistically-significantly inferior configurations. We name the method Bootstrap Corrected with Early Dropping CV (BCED-CV) that is both efficient and provides accurate performance estimates.

2016

  • V. Lagani, S. Triantafillou, G. Ball, J. Tegner, and I. Tsamardinos, «Probabilistic computational causal discovery for systems biology,» Uncertainty in Biology, vol. 17, pp. 33-73, 2016.
    [Summary]

  • G. Athineou, G. Papoutsoglou, S. Triantafullou, I. Basdekis, V. Lagani, and I. Tsamardinos, «SCENERY: a Web-Based Application for Network Reconstruction and Visualization of Cytometry Data.,» Accepted for publication on the 10th International Conference on Practical Applications of Computational Biology & Bioinformatics (PACBB 2016)., 2016.
    [Summary]

    Cytometry techniques allow to quantify morphological characteristics and protein abundances at a single-cell level. Data collected with these techniques can be used for addressing the fascinating, yet challenging problem of reconstructing the network of protein interactions forming signaling pathways and governing cell biological mechanisms. Network reconstruction is an established and well studied problem in the machine learning and data mining fields, with several algorithms already available. In this paper, we present the first web-oriented application, SCENERY, that allows scientists to rapidly apply state-of-the-art network-reconstruction methods on cytometry data. SCENERY comes with an easy-to-use user interface, a modular architecture, and advanced visualization functions. The functionalities of the application are illustrated on data from a publicly available immunology experiment.

  • O. D. Roe, M. Markaki, R. Mjelle, Pål. Sætrom, I. Tsamardinos, and V. Lagani, «Serum microRNAs/enriched pathways in lung cancer 1-4 years before diagnosis: A pilot study from the HUNT Biobank, Norway.,» in 2016 ASCO Annual Meeting, 2016.
    [Summary]

  • [DOI] V. Lagani, A. D. Karozou, D. Gomez-Cabrero, G. Silberberg, and I. Tsamardinos, «A comparative evaluation of data-merging and meta-analysis methods for reconstructing gene-gene interactions,» BMC Bioinformatics, vol. 17, iss. S5, p. 194, 2016.
    [Summary]

    We address the problem of integratively analyzing multiple gene expression, microarray datasets in order to reconstruct gene-gene interaction networks. Integrating multiple datasets is generally believed to provide increased statistical power and to lead to a better characterization of the system under study. However, the presence of systematic variation across different studies makes network reverse-engineering tasks particularly challenging. We contrast two approaches that have been frequently used in the literature for addressing systematic biases: meta-analysis methods, which first calculate opportune statistics on single datasets and successively summarize them, and data-merging methods, which directly analyze the pooled data after removing eventual biases. This comparative evaluation is performed on both synthetic and real data, the latter consisting of two manually curated microarray compendia comprising several E. coli and Yeast studies, respectively. Furthermore, the reconstruction of the regulatory network of the transcription factor Ikaros in human Peripheral Blood Mononuclear Cells (PBMCs) is presented as a case-study.

  • [DOI] A. I. Robles, K. Standahl Olsen, D. W. Tsui, V. Georgoulias, J. Creaney, K. Dobra, M. Vyberg, N. Minato, R. A. Anders, A. Børresen‑Dale, J. Zhou, Pål. Saetrom, B. Schnack Nielsen, M. B. Kirschner, H. E. Krokan, V. Papadimitrakopoulou, I. Tsamardinos, and O. D. Røe, «Excerpts from the 1st international NTNU symposium on current and future clinical biomarkers of cancer: innovation and implementation, June 16th and 17th 2016, Trondheim, Norway,» Journal of Translational Medicine, vol. 14, 2016.
    [Summary]

    The goal of biomarker research is to identify clinically valid markers. Despite decades of research there has been disap‑ pointingly few molecules or techniques that are in use today. The » 1st International NTNU Symposium on Current and Future Clinical Biomarkers of Cancer: Innovation and Implementation » , was held June 16th and 17th 2016, at the Knowledge Center of the St. Olavs Hospital in Trondheim, Norway, under the auspices of the Norwegian University of Science and Technology (NTNU) and the HUNT biobank and research center. The Symposium attracted approximately 100 attendees and invited speakers from 12 countries and 4 continents. In this Symposium original research and over‑ views on diagnostic, predictive and prognostic cancer biomarkers in serum, plasma, urine, pleural fluid and tumor, circulating tumor cells and bioinformatics as well as how to implement biomarkers in clinical trials were presented. Senior researchers and young investigators presented, reviewed and vividly discussed important new developments in the field of clinical biomarkers of cancer, with the goal of accelerating biomarker research and implementation. The excerpts of this symposium aim to give a cutting‑edge overview and insight on some highly important aspects of clinical cancer biomarkers to‑date to connect molecular innovation with clinical implementation to eventually improve patient care.

  • [DOI] N. Karathanasis, I. Tsamardinos, and V. Lagani, «omicsNPC: applying the NonParametric Combination methodology to the integrative analysis of heterogeneous omics data,» PloS one, p. e0165545, 2016.
    [Summary]

  • A. Roumpelaki, G. Borboudakis, S. Triantafillou, and I. Tsamardinos, «Marginal causal consistency in constraint-based causal learning,» in Uncertainty in Artificial Intelligence (UAI), 2016.
    [Summary]

    Maximal Ancestral Graphs (MAGs) are proba-bilistic graphical models that can model the dis-tribution and causal properties of a set of vari-ables in the presence of latent confounders. They are closed under marginalization. Invariant pair-wise features of a class of Markov equivalent MAGs can be learnt from observational data sets using the FCI algorithm and its variations (such as conservative FCI and order independent FCI). We investigate the consistency of causal features (causal ancestry relations) obtained by FCI in different marginals of a single data set. In prin-ciple, the causal relationships identified by FCI on a data set D measuring a set of variables V should not conflict the output of FCI on marginal data sets including only subsets of V. In prac-tice, however, FCI is prone to error propagation, and running FCI in different marginals results in inconsistent causal predictions. We introduce the term of marginal causal consistency to de-note the consistency of causal relationships when learning marginal distributions, and investigate the marginal causal consistency of different FCI variations.Results indicate that marginal causal consistency varies for different algorithms, and is also sensitive to network density and marginal size.

  • S. Triantafillou and I. Tsamardinos, «Score based vs constraint based causal learning in the presence of confounders,» in Uncertainty in Artificial Intelligence (UAI), 2016.
    [Summary]

  • [DOI] P. Charonyktakis, M. Plakia, I. Tsamardinos, and M. Papadopouli, «On user-centric modular QoE prediction for voip based on machine-learning algorithms,» IEEE Transactions on Mobile Computing, 2016.
    [Summary]

    The impact of the network performance on the quality of experience (QoE) for various services is not well-understood. Assessing the impact of different network and channel conditions on the user experience is important for improving the telecommunication services. The QoE for various wireless services including VoIP, video streaming, and web browsing, has been in the epicenter of recent networking activities. The majority of such efforts aim to characterize the user experience, analyzing various types of measurements often in an aggregate manner. This paper proposes the MLQoE, a modular algorithm for user-centric QoE prediction. The MLQoE employs multiple machine learning (ML) algorithms, namely, Artificial Neural Networks, Support Vector Regression machines, Decision Trees, and Gaussian Naive Bayes classifiers, and tunes their hyper-parameters. It uses the Nested Cross Validation (nested CV) protocol for selecting the best classifier and the corresponding best hyper-parameter values and estimates the performance of the final model. The MLQoE is conservative in the performance estimation despite multiple induction of models. The MLQoE is modular, in that, it can be easily extended to include other ML algorithms. The MLQoE selects the ML algorithm that exhibits the best performance and its parameters automatically given the dataset used as input. It uses empirical measurements based on network metrics (e.g., packet loss, delay, and packet interarrival) and subjective opinion scores reported by actual users. This paper extensively evaluates the MLQoE using three unidirectional datasets containing VoIP calls over wireless networks under various network conditions and feedback from subjects (collected in field studies). Moreover, it performs a preliminary analysis to assess the generality of our methodology using bidirectional VoIP and video traces. The MLQoE outperforms several state-of-the-art algorithms, resulting in fairly accurate predictions.

  • G. Borboudakis and I. Tsamardinos, «Towards Robust and Versatile Causal Discovery forBusiness Applications,» in Accepted at the 22nd SIGKDD Conference on Knowledge Discovery and Data Mining (KDD 2016), 2016.
    [Summary]

  • [DOI] J. Goveia, A. Pircher, L. Conradi, J. Kalucka, V. Lagani, M. Dewerchin, G. Eelen, R. J. DeBerardinis, I. D. Wilson, and P. Carmeliet, «Meta-analysis of clinical metabolic profiling studies in cancer: challenges and opportunities,» EMBO Molecular Medicine, 2016.
    [Summary]

2015

  • V. Lagani, F. Chiarugi, S. Thomson, J. Fursse, E. Lakasing, R. W. Jones, and I. Tsamardinos, «Development and validation of risk assessment models for diabetes-related complications based on the DCCT/EDIC data,» Journal of Diabetes and its Complications 2015., vol. 29, iss. 4, pp. 479-487, 2015.
    [Summary]

  • I. Tsamardinos, M. Tsagris, and V. Lagani, «Feature selection for longitudinal data,» Proceedings of the 10th conference of the Hellenic Society for Computational Biology & Bioinformatics (HSCBB15), vol. 6, iss. 1, pp. 10-11, 2015.
    [Summary]

  • A. Alexandridis, G. Borboudakis, and A. Mouchtaris, «Addressing the data-association problem for multiple sound source localization using DOA data estimates,» 23rd European Signal Processing Conference (EUSIPCO), pp. 1551-1555, 2015.
    [Summary]

    In this paper, we consider the data association problem that arises when localizing multiple sound sources using direction of arrival (DOA) estimates from multiple microphone arrays. In such a sce- nario, the association of the DOAs across the arrays that correspond to the same source is unknown and must be found for accurate lo- calization. We present an association algorithm that finds the correct DOA association to the sources based on features extracted for each source that we propose. Our method results in high association and localization accuracy in scenarios with missed detections, reverber- ation, and noise and outperforms other recently proposed methods.

  • N. Karathanasis, I. Tsamardinos, and P. Poirazi, «MiRduplexSVM: A high-Performing miRNA-duplex prediction and evaluation methodology,» PloS one, vol. 10, iss. 5, 2015.
    [Summary]

  • I. Tsamardinos, A. Rakhshani, and V. Lagani, «Performance-estimation properties of cross-validation based protocols with simultaneous hyper-parameter optimization,» International Journal on Artificial Intelligence Tools, vol. 8445, pp. 1-14, 2015.
    [Summary]

  • V. Lagani, F. Chiarugi, D. Manousos, V. Verma, J. Fursse, M. Kostas, and I. Tsamardinos, «Realization of a service for the long-term risk assessment of diabetes-related complications,» Journal of Diabetes and its Complications, vol. 29, iss. 5, pp. 691-698, 2015.
    [Summary]

2014

  • K. Kerkentzes, V. Lagani, I. Tsamardinos, M. Vyberg, and O. Røe, «Hidden treasures in “ancient” microarrays: gene-expression portrays biology and potential resistance pathways of major lung cancer subtypes and normal,» Frontiers 2014, vol. 4, iss. 251, 2014.
    [Summary]

  • G. T. Huang, I. Tsamardinos, V. Raghu, N. Kaminski, and P. V. Benos, «T-ReCS: stable selection of dynamically formed groups of features with application to prediction of clinical outcomes.,» Pacific Symposium on Biocomputing (PSB), pp. 431-442, 2014.
    [Summary]

  • S. Triantafillou and I. Tsamardinos, «Constraint-based Causal Discovery from Multiple Interventions over Overlapping Variable Sets,» Journal of Machine Learning Research, vol. 16, pp. 1-47, 2014.
    [Summary]

    Scientific practice typically involves repeatedly studying a system, each time trying to unravel a different perspective. In each study, the scientist may take measurements under different experimental conditions (interventions, manipulations, perturbations) and measure different sets of quantities (variables). The result is a collection of heterogeneous data sets coming from different data distributions. In this work, we present algorithm COmbINE, which accepts a collection of data sets over overlapping variable sets under different experimental conditions; COmbINE then outputs a summary of all causal models indicating the invariant and variant structural characteristics of all models that simultaneously fit all of the input data sets. COmbINE converts estimated dependencies and independencies in the data into path constraints on the data-generating causal model and encodes them as a SAT instance. The algorithm is sound and complete in the sample limit. To account for conflicting constraints arising from statistical errors, we introduce a general method for sorting constraints in order of confidence, computed as a function of their corresponding p-values. In our empirical evaluation, COmbINE outperforms in terms of efficiency the only pre-existing similar algorithm; the latter additionally admits feedback cycles, but does not admit conflicting constraints which hinders the applicability on real data. As a proof-of-concept, COmbINE is employed to co-analyze 4 real, mass-cytometry data sets measuring phosphorylated protein concentrations of overlapping protein sets under 3 different interventions.

  • C. Papagiannopoulou, G. Tsoumakas, and I. Tsamardinos, «Discovering and Exploiting Entailment Relationships in Multi-Label Learning,» ACM SIGKDD Conference on Knowledge Discovery and Data Mining 2015 (KDD), 2014.
    [Summary]

  • I. Tsamardinos, A. Rakhshani, and V. Lagani, «Performance-estimation properties of cross-validation-based protocols with simultaneous hyper-parameter optimization,» 8th Hellenic Conference on Artificial Intelligence (SETN 2014)., vol. 8445, pp. 1-14, 2014.
    [Summary]

  • N. Karathanasis, I. Tsamardinos, and P. Poirazi, «Don’t use a cannon to kill the miRNA mosquito,» Bioinformatics, vol. 30, iss. 7, pp. 1047-1048, 2014.
    [Summary]

  • S. Triantafilou, I. Tsamardinos, and A. Roumpelaki, «Learning Neighborhoods of High Confidence in Constraint-Based Causal Discovery.,» Springer 2014, vol. 8754, pp. 487-502, 2014.
    [Summary]

2013

  • V. Lagani, G. Kortas, and I. Tsamardinos, «Biomarker signature identification in “omics” data with multi-class outcome,» Computational and Structural Biotechnology Journal, 2013, 6(7), vol. 6, iss. 7, pp. 1-7, 2013.
    [Summary]

  • P. Hunter, T. Chapman, P. Coveney, B. de Bono, V. Diaz, J. Fenner, A. Frangi, P. Harris, R. Hose, P. Kohl, P. Lawford, K. McCormack, M. Mendes, S. Omholt, A. Quarteroni, N. Shublaq, J. Skår, K. Stroetmann, J. Tegner, S. Thomas, I. Tollis, I. Tsamardinos, J. van Beek, and M. Viceconti, «A vision and strategy for the virtual physiological human: 2012 update,» Interface Focus, vol. 3, iss. 2, pp. 1-27, 2013.
    [Summary]

  • N. Karathanasis, I. Tsamardinos, and P. Poirazi, «A bioinformatics approach for investigating the determinants of Drosha processing,» 13th IEEE International Conference on Bioinformatics and Bioengineering (IEEE BIBE 2013), pp. 1-4, 2013.
    [Summary]

  • I. Karakasilioti, I. Kamileri, G. Chatzinikolaou, T. Kosteas, E. Vergadi, A. R. Robinson, I. Tsamardinos, T. A. Rozgaja, S. Siakouli, C. Tsatsanis, L. J. Niedernhofer, and G. A. Garinis, «DNA damage triggers a chronic autoinflammatory response, leading to fat depletion in NER progeria,» Cell Metabolism, vol. 18, iss. 3, pp. 403-415, 2013.
    [Summary]

  • G. L. Papadopoulos, E. Karkoulia, I. Tsamardinos, C. Porcher, J. Ragoussis, J. Bungert, and J. Strouboulis, «GATA-1 genome-wide occupancy associates with distinct epigenetic profiles in mouse fetal liver erythropoiesis,» Nucl. Acids Res, vol. 41, iss. 9, pp. 4938-4948, 2013.
    [Summary]

  • V. Lagani, L. Koumakis, F. Chiarugi, E. Lakasing, and I. Tsamardinos, «A systematic review of predictive risk models for diabetes complications based on large scale clinical studies,» Journal of Diabetes and Its Complications, vol. 27, iss. 4, pp. 407-413, 2013.
    [Summary]

2012

  • G. Borboudakis and I. Tsamardinos, «Scoring and searching over Bayesian networks with causal and associative priors,» Uncertainty in Artificial Intelligence (UAI), 2012.
    [Summary]

  • S. Kleisarchaki, D. Kotzinos, I. Tsamardinos, and V. Christophides, «A Methodological Framework for Statistical Analysis of Social Text Streams,» International Workshop on Information Search, Integration and Personalization (ISIP 2012), vol. 146, pp. 101-110, 2012.
    [Summary]

  • I. Tsamardinos, S. Triantafillou, and V. Lagani, «Towards integrative causal analysis of heterogeneous data sets and studies,» Journal of Machine Learning Research, vol. 13, iss. 1, pp. 1097-1157, 2012.
    [Summary]

  • T. I. Brown L.E. and D. Hardin, «To feature space and back: Identifying top-weighted features in polynomial support vector machine models,» Intelligent Data Analysis, vol. 16, iss. 4, pp. 551-579, 2012.
    [Summary]

  • I. Tsamardinos, G. Borboudakis, E. Christodoulou, and O. D. Røe, «Chemosensitivity Prediction of Tumours Based on Expression, miRNA, and Proteomics Data,» International Journal of Systems Biology and Biomedical Technologies (IJSBBT), vol. 1, iss. 2, pp. 1-19, 2012.
    [Summary]

  • G. Borboudakis and I. Tsamardinos, «Incorporating causal prior knowledge as path-constraints in bayesian networks and maximal ancestral graphs,» Proceedings of the 29th International Conference on Machine Learning, ICML 2012, pp. 1799-1806, 2012.
    [Summary]

  • V. Lagani, I. Tsamardinos, and S. Triantafillou, «Learning from Mixture of Experimental Data: A Constraint–Based Approach,» Artificial Intelligence: Theories and Applications: 7th Hellenic Conference on AI, vol. 7297, pp. 124-131, 2012.
    [Summary]

  • A. P. Armen, I. Tsamardinos, N. Karathanasis, and P. Poirazi, «SVM-based miRNA: MiRNA∗ duplex prediction,» IEEE 12th International Conference on Bioinformatics and Bioengineering, BIBE 2012, pp. 181-186, 2012.
    [Summary]

  • G. Borboudakis, S. Triantafillou, and I. Tsamardinos, «Tools and algorithms for causally interpreting directed edges in maximal ancestral graphs,» Sixth European Workshop on Probabilistic Graphical Models, (PGM 2012), 2012.
    [Summary]

  • S. Dato, A. Montesanto, V. Lagani, B. Jeune, K. Christensen, and G. Passarino, «Frailty phenotypes in the elderly based on cluster analysis: a longitudinal study of two Danish cohorts. Evidence for a genetic influence on frailty,» Age, vol. 34, iss. 3, pp. 571-582, 2012.
    [Summary]

  • S. Dato, M. Soerensen, A. Montesanto, V. Lagani, G. Passarino, K. Christensen, and L. Christiansen, «UCP3 polymorphisms, hand grip performance and survival at old age: Association analysis in two Danish middle aged and elderly cohorts,» Joint Meeting AGI-SIBV-SIGA, vol. 133, iss. 8, pp. 530-537, 2012.
    [Summary]

  • I. Tsamardinos, V. Lagani, and D. Pappas, «Discovering multiple, equivalent biomarker signatures.,» proceedings of the 7th conference of the Hellenic Society for Computational Biology & Bioinformatics, pp. 1-2, 2012.
    [Summary]

2011

  • G. Borboudakis, S. Triantafilou, V. Lagani, and I. Tsamardinos, «A constraint-based approach to incorporate prior knowledge in causal models.,» Proceedings of the European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN), 2011.
    [Summary]

  • A. Armen and I. Tsamardinos, «A unified approach to estimation and control of the False Discovery Rate in Bayesian network skeleton identification.,» Proceedings of the European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN), 2011.
    [Summary]

  • E. G. Christodoulou, O. D. Røe, A. Folarin, and I. Tsamardinos, «Information-Preserving Techniques Improve Chemosensitivity Prediction of Tumours Based on Expression Profiles,» 12th Engineering Applications of Neural Networks (EANN) / 7th Artificial Intelligence Applications and Innovations (AIAI) joint conferences, Workshop on Computational Intelligence Applications in Bioinformatics (CIAB 2011), vol. 363, pp. 453-462, 2011.
    [Summary]

  • L. Koumakis, F. Chiarugi, V. Lagani, and I. Tsamardinos, «Risk assessment models for diabetes complications: A survey of available online tools,» 2nd International ICST Conference on Wireless Mobile Communication and Healthcare (MobiHealth 2011), vol. 83, pp. 46-53, 2011.
    [Summary]

  • V. Lagani, V. Kontogiannis, P. Argyropaidas, and C. Chronaki, «Use of SMS for Tsunami Early Warnings at a Table Top Exercise,» 4th ICST International Conference on eHealth (eHealth 2011), vol. 91, pp. 76-79, 2011.
    [Summary]

  • C. Filippaki, G. Antoniou, and I. Tsamardinos, «Using constraint optimization for conflict resolution and detail control in activity recognition,» Second International Joint Conference on Ambient Intelligence 2011, vol. 7040, pp. 51-60, 2011.
    [Summary]

  • V. Lagani, I. Tsamardinos, M. Grammatikou, and G. Garinis, «A Genome-Wide Study of the Effect of Aging on Level-2 Gene-Ontology Categories in Mice Using Mixed Models,» The 5th International Workshop on Data Mining in Functional Genomics and Proteomics: Current Trends and Future Directions, ECML PKDD 2011, pp. 1-10, 2011.
    [Summary]

  • I. Tsamardinos, O. D. Roe, and V. Lagani, «Introducing Integrative Causal Analysis for Co-Analyzing Heterogeneous Studies with an Application to Methylation and Gene Expression Mesothelioma Cancer Data.,» 6th Conference of the Hellenic Society for Computational Biology and Bioinformatics (HSCBB11), 2011.
    [Summary]

2010

  • S. Triantafullou, I. Tsamardinos, and I. G. Tollis, «Learning causal structure from overlapping variable sets,» Proceedings of The Thirteenth International Conference on Artificial Intelligence and Statistics (AISTATS), vol. 9, pp. 860-867, 2010.
    [Summary]

  • I. Tsamardinos and G. Borboudakis, «Permutation testing improves Bayesian network learning,» Proceedings of the 2010 European Conference on Machine Learning and Knowledge Discovery in Databases: Part III, vol. 6323, pp. 322-337, 2010.
    [Summary]

  • V. Lagani and I. Tsamardinos, «Structure-based variable selection for survival data,» Bioinformatics, vol. 26, iss. 15, pp. 1887-1894, 2010.
    [Summary]

  • K. Gkirtzou, I. Tsamardinos, P. Tsakalides, and P. Poirazi, «MatureBayes: a probabilistic algorithm for identifying the mature miRNA within novel precursors,» PloS one, vol. 5, iss. 8, pp. 1-14, 2010.
    [Summary]

  • P. Hunter, P. V. Coveney, B. de Bono, V. Diaz, J. Fenner, A. F. Frangi, P. Harris, R. Hose, P. Kohl, P. Lawford, K. McCormack, M. Mendes, S. Omholt, A. Quarteroni, J. Skar, J. Tegner, S. Randall, I. G. Tollis, I. Tsamardinos, and M. van Beek, «A vision and strategy for the virtual physiological human in 2010 and beyond,» Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, vol. 368, iss. 1920, pp. 2595-2614, 2010.
    [Summary]

  • F. Chiarugi, D. Emmanouilidou, and I. Tsamardinos, «The morphological classification of heartbeats as dominant and non-dominant in ECG signals,» Physiological Measurement, vol. 31, iss. 5, pp. 611-631, 2010.
    [Summary]

  • C. F. Aliferis, A. R. Statnikov, I. Tsamardinos, S. Mani, and X. D. Koutsoukos, «Local causal and markov blanket induction for causal discovery and feature selection for classification part i: Analysis and extensions,» The Journal of Machine Learning Research, vol. 11, pp. 171-234, 2010.
    [Summary]

    In part I of this work we introduced and evaluated the Generalized Local Learning (GLL) framework for producing local causal and Markov blanket induction algorithms. In the present second part we analyze the behavior of GLL algorithms and provide extensions to the core methods. Specifically, we investigate the empirical convergence of GLL to the true local neighborhood as a function of sample size. Moreover, we study how predictivity improves with increasing sample size. Then we investigate how sensitive are the algorithms to multiple statistical testing, especially in the presence of many irrelevant features. Next we discuss the role of the algorithm parameters and also show that Markov blanket and causal graph concepts can be used to understand deviations from optimality of state-of-the-art non-causal algorithms. The present paper also introduces the following extensions to the core GLL framework: parallel and distributed versions of GLL algorithms, versions with false discovery rate control, strategies for constructing novel heuristics for specific domains, and divide-and-conquer local-to-global learning (LGL) strategies. We test the generality of the LGL approach by deriving a novel LGL-based algorithm that compares favorably to the state-of-the-art global learning algorithms. In addition, we investigate the use of non-causal feature selection methods to facilitate global learning. Open problems and future research paths related to local and local-to-global causal learning are discussed.

  • C. F. Aliferis, A. R. Statnikov, I. Tsamardinos, S. Mani, and X. D. Koutsoukos, «Local causal and markov blanket induction for causal discovery and feature selection for classification part ii: Analysis and extensions,» The Journal of Machine Learning Research, vol. 11, pp. 235-284, 2010.
    [Summary]

    In part I of this work we introduced and evaluated the Generalized Local Learning (GLL) framework for producing local causal and Markov blanket induction algorithms. In the present second part we analyze the behavior of GLL algorithms and provide extensions to the core methods. Specifically, we investigate the empirical convergence of GLL to the true local neighborhood as a function of sample size. Moreover, we study how predictivity improves with increasing sample size. Then we investigate how sensitive are the algorithms to multiple statistical testing, especially in the presence of many irrelevant features. Next we discuss the role of the algorithm parameters and also show that Markov blanket and causal graph concepts can be used to understand deviations from optimality of state-of-the-art non-causal algorithms. The present paper also introduces the following extensions to the core GLL framework: parallel and distributed versions of GLL algorithms, versions with false discovery rate control, strategies for constructing novel heuristics for specific domains, and divide-and-conquer local-to-global learning (LGL) strategies. We test the generality of the LGL approach by deriving a novel LGL-based algorithm that compares favorably to the state-of-the-art global learning algorithms. In addition, we investigate the use of non-causal feature selection methods to facilitate global learning. Open problems and future research paths related to local and local-to-global causal learning are discussed.

2009

  • C. F. Aliferis, A. R. Statnikov, I. Tsamardinos, J. S. Schildcrout, B. E. Shepherd, and J. F. E. Harrell, «Factors influencing the statistical power of complex data analysis protocols for molecular signature development from microarray data,» PloS one, vol. 4, iss. 3, pp. 1-7, 2009.
    [Summary]

  • I. Tsamardinos and A. Mariglis, «Multi-source causal analysis: Learning Bayesian networks from multiple datasets,» Artificial Intelligence Applications and Innovations III, Proceedings of the 5TH IFIP Conference on Artificial Intelligence Applications and Innovations, vol. 296, pp. 479-490, 2009.
    [Summary]

  • I. Tsamardinos and S. Triantafillou, «The possibility of integrative causal analysis: Learning from different datasets and studies,» Journal of Engineering Intelligent Systems, vol. 17, iss. 2-3, pp. 81-93, 2009.
    [Summary]

2008

  • L. Brown and I. Tsamardinos, «A Strategy for Making Predictions Under Manipulation.,» In JMLR: Workshop and Conference Proceedings, vol. 3, pp. 35-52, 2008.
    [Summary]

  • I. Tsamardinos and L. Brown, «Bounding the False Discovery Rate in Local Bayesian Network Learning.,» Proceedings of the Twenty-Third AAAI Conference on Artificial Intelligence, pp. 1100-1105, 2008.
    [Summary]

  • F. Chiarugi, D. Emmanouilidou, I. Tsamardinos, and I. G. Tollis, «Morphological classification of heartbeats using similarity features and a two-phase decision tree,» Computers in Cardiology, pp. 849-852, 2008.
    [Summary]

2006

  • C. Aliferis, A. Statnikov, and I. Tsamardinos, «Challenges in the analysis of mass-throughput data: a technical commentary from the statistical machine learning perspective,» Cancer Informatics, vol. 2, pp. 133-162, 2006.
    [Summary]

  • I. Tsamardinos, A. Statnikov, L. Brown, and C. Aliferis, «Generating Realistic Large Bayesian Networks by Tiling.,» In The 19th International FLAIRS Conference, pp. 592-597, 2006.
    [Summary]

  • I. Tsamardinos, L. Brown, and C. Aliferis, «The max-min hill-climbing Bayesian network structure learning algorithm,» Machine learning, vol. 65, iss. 1, pp. 31-78, 2006.
    [Summary]

2005

  • A. Statnikov, I. Tsamardinos, Y. Dosbayev, and C. F. Aliferis, «GEMS: a system for automated cancer diagnosis and biomarker discovery from microarray gene expression data,» International journal of medical informatics, vol. 74, iss. 7-8, pp. 491-504, 2005.
    [Summary]

  • L. Brown, I. Tsamardinos, and C. Aliferis, «A comparison of novel and state-of-the-art polynomial Bayesian network learning algorithms,» In Proceedings of the national conference on artificial intelligence, vol. 20, iss. 2, pp. 739-745, 2005.
    [Summary]

  • Y. Aphinyanaphongs, I. Tsamardinos, A. Statnikov, D. Hardin, and C. F. Aliferis, «Text categorization models for high-quality article retrieval in internal medicine,» Journal of the American Medical Informatics Association, vol. 12, iss. 2, pp. 207-216, 2005.
    [Summary]

  • A. Statnikov, C. F. Aliferis, I. Tsamardinos, D. Hardin, and S. Levy, «A comprehensive evaluation of multicategory classification methods for microarray gene expression cancer diagnosis,» Bioinformatics, vol. 21, iss. 5, pp. 631-643, 2005.
    [Summary]

  • M. Pollack and I. Tsamardinos, «Efficiently dispatching plans encoded as simple temporal problems,» Intelligent Techniques for Planning, pp. 296-319, 2005.
    [Summary]

2004

  • D. Hardin, I. Tsamardinos, and C. Aliferis, «A theoretical characterization of linear SVM-based feature selection,» The Twenty-First International Conference on Machine Learning (ICML 2004), pp. 48-55, 2004.
    [Summary]

  • L. Brown, I. Tsamardinos, and C. Aliferis, «A novel algorithm for scalable and accurate Bayesian network learning,» in Proceedings of 11th World Congress in Medical Informatics (MEDINFO ’04), vol. 107, iss. 1, pp. 711-715, 2004.
    [Summary]

  • A. Statnikov, C. Aliferis, and I. Tsamardinos, «Methods for multi-category cancer diagnosis from gene expression data: a comprehensive evaluation to inform decision support system development,» in Proceedings of the 11th World Congress in Medical Informatics (MEDINFO ’04), vol. 107, iss. 2, pp. 813-817, 2004.
    [Summary]

2003

  • L. Frey, D. Fisher, I. Tsamardinos, C. F. Aliferis, and A. Statnikov, «Identifying Markov blankets with decision tree induction,» The Third IEEE International Conference on Data Mining (ICDM’03), pp. 59-66, 2003.
    [Summary]

  • C. F. Aliferis, I. Tsamardinos, and A. Statnikov, «HITON: a novel Markov Blanket algorithm for optimal variable selection,» the American Medical Informatics Association meeting 2003 (AMIA 2003), pp. 21-25, 2003.
    [Summary]

  • I. Tsamardinos, C. Aliferis, and A. Statnikov, «Time and sample efficient discovery of Markov blankets and direct causal relations,» The Ninth ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD 2003), pp. 673-678, 2003.
    [Summary]

  • I. Tsamardinos, M. E. Pollack, and S. Ramakrishnan, «Assessing the probability of legal execution of plans with temporal uncertainty,» ICAPS03 Workshop on Planning under Un-certainty and Incomplete Information, pp. 110-118, 2003.
    [Summary]

  • M. E. Pollack, L. Brown, D. Colbry, C. E. McCarthy, C. Orosz, B. Peintner, S. Ramakrishnan, and I. Tsamardinos, «Autominder: An intelligent cognitive orthotic system for people with memory impairment,» Robotics and Autonomous Systems, vol. 44, iss. 3-4, pp. 273-282, 2003.
    [Summary]

  • I. Tsamardinos and M. Pollack, «Efficient solution techniques for disjunctive temporal reasoning problems,» Artificial Intelligence, vol. 151, iss. 1-2, pp. 43-89, 2003.
    [Summary]

  • I. Tsamardinos, T. Vidal, and M. Pollack, «CTP: A new constraint-based formalism for conditional, temporal planning,» Special Issue on Planning of Constraints Journal, vol. 8, iss. 4, pp. 365-388, 2003.
    [Summary]

  • C. Aliferis, I. Tsamardinos, A. Statnikov, and L. Brown, «Causal Explorer: A Causal Probabilistic Network Learning Toolkit for Biomedical Discovery.,» International Conference on Mathematics and Engineering Techniques in Medicine and Biolog-ical Sciences (METMBS ’03), pp. 371-376, 2003.
    [Summary]

  • C. F. Aliferis, I. Tsamardinos, P. Massion, A. Statnikov, and D. Hardin, «Why Classification Models Using Array Gene Expression Data Perform So Well: A Preliminary Investigation of Explanatory Factors.,» International Conference on Mathematics and Engineering Techniques in Medicine and Biological Sciences (METMBS ’03), pp. 47-53, 2003.
    [Summary]

  • I. Tsamardinos, C. Aliferis, A. Statnikov, and E. Statnikov, «Algorithms for Large Scale Markov Blanket Discovery.,» FLAIRS conference, pp. 376-381, 2003.
    [Summary]

  • I. Tsamardinos and C. Aliferis, «Towards Principled Feature Selection: Relevancy, Filters and Wrappers.,» in Ninth International Workshop on Artificial Intelligence and Statistics, AISTAT, 2003.
    [Summary]

2002

  • I. Tsamardinos, «A probabilistic approach to robust execution of temporal plans with uncertainty,» Proceedings of the 2nd Greek National Conference on Artificial Intelligence, vol. 2308, pp. 97-108, 2002.
    [Summary]

  • M. E. Pollack, C. E. McCarthy, S. Ramakrishnan, and I. Tsamardinos, «Execution-time plan management for a cognitive orthotic system,» Plan-Based Control of Robotic Agents, vol. 2466, pp. 179-192, 2002.
    [Summary]

  • M. E. Pollack, C. E. McCarthy, S. Ramakrishnan, I. Tsamardinos, L. Brown, S. Carrion, D. Colbry, C. Orosz, and B. Peintner, «Autominder: A planning, monitoring, and reminding assistive agent,» Proceedings of the 7th International Conference on Intelligent Autonomous Systems (IAS), 2002.
    [Summary]

  • A. Berfield, P. K. Chrysanthis, I. Tsamardinos, M. E. Pollack, and S. Banerjee, «A scheme for integrating e-services in establishing virtual enterprises,» 12th International Workshop on Research Issues on Data Engineering (RIDE-02), pp. 134-142, 2002.
    [Summary]

  • C. F. Aliferis, I. Tsamardinos, P. Mansion, A. Statnikov, and D. Hardin, «Machine learning models for lung cancer classification using array comparative genomic hybridization.,» 16th International FLAIRS Conference, pp. 67-71, 2002.
    [Summary]

    Array CGH is a recently introduced technology that measures changes in the gene copy number of hundreds of genes in a single experiment. The primary goal of this study was to develop machine learning models that classify non-small Lung Cancers according to histopathology types and to compare several machine learning methods in this learning task. DNA from tumors of 37 patients (21 squamous carcinomas, and 16 adenocarcinomas) were extracted and hybridized onto a 452 BAC clone array. The following algorithms were used: KNN, Decision Tree Induction, Support Vector Machines and Feed-Forward Neural Networks. Performance was measured via leave-one-out classification accuracy. The best multi-gene model found had a leave-one-out accuracy of 89.2\%. Decision Trees performed poorer than the other methods in this learning task and dataset. We conclude that gene copy numbers as measured by array CGH are, collectively, an excellent indicator of histological subtype. Several interesting research directions are discussed.

2001

  • I. Tsamardinos, M. Pollack, and P. Ganchev, «Flexible dispatch of disjunctive plans,» Proceedings of Sixth European Conference on Planning 2001 (ECP-01), pp. 417-422, 2001.
    [Summary]

2000

  • I. Tsamardinos, M. Pollack, and J. Horty, «Merging Plans with Quantitative Temporal Constraints, Temporally Extended Actions, and Conditional Branches.,» Proceedings of the 5th International Conference on AI Planning and Scheduling (AIPS 2000), Breckenridge, CO, April, 2000, pp. 264-272, 2000.
    [Summary]

1999

  • M. Pollack, I. Tsamardinos, and J. Horty, «Adjustable autonomy for a plan management agent,» Proceedings of the 1999 AAAI Spring Symposium on Adjustable Autonomy, 1999.
    [Summary]

1998

  • I. Tsamardinos, N. Muscettola, and P. Morris, «Fast transformation of temporal plans for efficient execution,» Proceedings of the 15th National Conference on Artificial Intelli-gence (AAAI’98), pp. 254-261, 1998.
    [Summary]

  • N. Muscettola, P. Morris, and I. Tsamardinos, «Reformulating temporal plans for efficient execution,» Proceedings of the 6th Conference Principles of Knowledge Represen-tation and Reasoning (KR), pp. 444-452, 1998.
    [Summary]

  • C. Bicchieri, M. Pollack, C. Rovelli, and I. Tsamardinos, «Bicchieri-The\_Potential\_for\_the\_Evolution\_of\_Co-operatin.pdf,» International Journal of Computer-Human Systems, vol. 48, iss. 1, pp. 9-29, 1998.
    [Summary]

1995

  • S. Orphanoudakis, M. Tsiknakis, C. Chronaki, S. Kostomanolakis, M. Zikos, and I. Tsamardinos, «Development of an Integrated Image Management and Communication System on Crete,» Proceedings of Computed Aided Radiology ’95, pp. 481-487, 1995.
    [Summary]