@article {1602, title = {Column generation-based prototype learning for optimizing area under the receiver operating characteristic curve}, journal = {European Journal of Operational Research}, volume = {314}, year = {2024}, pages = {297{\textendash}307}, author = {Erhan C Ozcan and Berk G{\"o}rg{\"u}l{\"u} and Mustafa Gokce Baydogan} } @article {1538, title = {Multiple instance classification via quadratic programming}, journal = {Journal of Global Optimization}, year = {2022}, pages = {1{\textendash}32}, author = {Emel {\c S}eyma K{\"u}{\c c}{\"u}ka{\c s}c{\i} and Mustafa Gokce Baydogan and Z C Ta{\c s}k{\i}n} } @article {1536, title = {nTreeClus: A tree-based sequence encoder for clustering categorical series}, journal = {Neurocomputing}, volume = {494}, year = {2022}, pages = {224-241}, abstract = {

The overwhelming presence of categorical/sequential data in diverse domains emphasizes the importance of sequence mining. The challenging nature of sequences proves the need for continuing research to find a more accurate and faster approach providing a better understanding of their (dis) similarities. This paper proposes a new Model-based approach for clustering sequence data, namely nTreeClus. The proposed method deploys Tree-based Learners, k-mers, and autoregressive models for categorical time series, culminating with a novel numerical representation of the categorical sequences. Adopting this new representation, we cluster sequences, considering the inherent patterns in categorical time series. Accordingly, the model showed robustness to its parameter. Under different simulated scenarios, nTreeClus improved the baseline methods for various internal and external cluster validation metrics for up to 10.7\% and 2.7\%, respectively. The empirical evaluation using synthetic and real datasets, protein sequences, and categorical time series showed that nTreeClus is competitive or superior to most state-of-the-art algorithms.

}, keywords = {Categorical time series, Model-based clustering, Pattern recognition, Sequence mining, Tree-based learning}, issn = {0925-2312}, doi = {https://doi.org/10.1016/j.neucom.2022.04.076}, url = {https://www.sciencedirect.com/science/article/pii/S0925231222004611}, author = {Hadi Jahanshahi and Mustafa Gokce Baydogan} } @article {1537, title = {Semi-supervised extensions of multi-task tree ensembles}, journal = {Pattern Recognition}, volume = {123}, year = {2022}, pages = {108393}, abstract = {

Scale inconsistency is a widely encountered issue in multi-output learning problems. Specifically, target sets with multiple real valued or a mixture of categorical and real valued variables require addressing the scale differences to obtain predictive models with sufficiently good performance. Data transformation techniques are often employed to solve that problem. However, these operations are susceptible to different shortcomings such as changing the statistical properties of the data and increase the computational burden. Scale differences also pose problem in semi-supervised learning (SSL) models as they require processing of unsupervised information where distance measures are commonly employed. Classical distance metrics can be criticized as they lose efficiency when variables exhibit type or scale differences, too. Besides, in higher dimensions distance metrics cause problems due to loss of discriminative power. This paper introduces alternative semi-supervised tree-based strategies that are robust to scale differences both in terms of feature and target variables. We propose use of a scale-invariant proximity measure by means of tree-based ensembles to preserve the original characteristics of the data. We update classical tree derivation procedure to a multi-criteria form to resolve scale inconsistencies. We define proximity based clustering indicators and extend the supervised model with unsupervised criteria. Our experiments show that proposed method significantly outperforms its benchmark learning model that is predictive clustering trees.

}, keywords = {Ensemble learning, Multi-objective trees, Multi-task learning, Semi-supervised learning, Totally randomized trees}, issn = {0031-3203}, doi = {https://doi.org/10.1016/j.patcog.2021.108393}, url = {https://www.sciencedirect.com/science/article/pii/S0031320321005549}, author = {Esra Ad{\i}yeke and Mustafa Gokce Baydogan} } @article {1423, title = {An ensemble-based semi-supervised feature ranking for multi-target regression problems}, journal = {Pattern Recognition Letters}, volume = {148}, year = {2021}, pages = {36-42}, abstract = {

This study focuses on semi-supervised feature ranking (FR) applications for multi-target regression problems (MTR). As MTRs require prediction of several targets, we use a learning model that includes target interrelations via multi-objective trees. In processing the features for a semi-supervised learning model, transformation or scaling operations are usually required. To resolve this issue, we create a dissimilarity matrix via totally randomized trees to process the unsupervised information. Besides, we treat the split score function as a vector to make it suitable for considering each criterion regardless of their scales. We propose a semi-supervised FR scheme embedded to multi-objective trees that takes into account target and feature contributions simultaneously. Proposed FR score is compared with the state-of-the-art multi-target FR strategies via statistical analyses. Experimental studies show that proposed score significantly improves the performance of a recent tree-based and competitive multi-target learning model, i.e. predictive clustering trees. In addition, proposed approach outperforms its benchmarks when the available labelled data increase.

}, keywords = {Feature ranking, Multi-target regression, Semi-supervised learning}, issn = {0167-8655}, doi = {https://doi.org/10.1016/j.patrec.2021.04.025}, url = {https://www.sciencedirect.com/science/article/pii/S0167865521001641}, author = {Esra Ad{\i}yeke and Mustafa Gokce Baydogan} } @article {1420, title = {Explainable boosted linear regression for time series forecasting}, journal = {Pattern Recognition}, volume = {120}, year = {2021}, pages = {108144}, abstract = {

Time series forecasting involves collecting and analyzing past observations to develop a model to extrapolate such observations into the future. Forecasting of future events is important in many fields to support decision making as it contributes to reducing the future uncertainty. We propose explainable boosted linear regression (EBLR) algorithm for time series forecasting, which is an iterative method that starts with a base model, and explains the model{\textquoteright}s errors through regression trees. At each iteration, the path leading to highest error is added as a new variable to the base model. In this regard, our approach can be considered as an improvement over general time series models since it enables incorporating nonlinear features by residual explanation. More importantly, use of the single rule that contributes to the error most enables access to interpretable results. The proposed approach extends to probabilistic forecasting through generating prediction intervals based on the empirical error distribution. We conduct a detailed numerical study with EBLR and compare against various other approaches. We observe that EBLR substantially improves the base model performance through extracted features, and provide a comparable performance to other well established approaches. The interpretability of the model predictions and high predictive accuracy of EBLR makes it a promising method for time series forecasting.

}, keywords = {ARIMA, Decision trees, Linear regression, Probabilistic forecasting, Time series regression}, issn = {0031-3203}, doi = {https://doi.org/10.1016/j.patcog.2021.108144}, url = {https://www.sciencedirect.com/science/article/pii/S0031320321003319}, author = {Igor Ilic and Berk G{\"o}rg{\"u}l{\"u} and Mucahit Cevik and Mustafa Gokce Baydogan} } @article {1535, title = {Learning prototypes for multiple instance learning.}, journal = {Turkish Journal of Electrical Engineering \& Computer Sciences}, volume = {29}, year = {2021}, author = {{\"O}zg{\"u}r Emre Sivrikaya and Mert Y{\"u}kselg{\"o}n{\"u}l and Mustafa Gokce Baydogan} } @article {1424, title = {A linear programming approach to multiple instance learning}, journal = {Turkish Journal of Electrical Engineering \& Computer Sciences}, volume = {29}, year = {2021}, pages = {2186{\textendash}2201}, author = {Emel {\c S}eyma K{\"u}{\c c}{\"u}ka{\c s}c{\i} and Mustafa Gokce Baydogan and Z C Ta{\c s}k{\i}n} } @article {1422, title = {A new feature-based time series classification method by using scale-space extrema}, journal = {Engineering Science and Technology, an International Journal}, year = {2021}, abstract = {

Time series data mining has received significant attention over the past decade, and many approaches have focused on classification tasks where the goal is to define the label of a test time series, given labeled training data. Time series classification approaches can be broadly grouped into two categories as instance-based and feature-based methods. Instance-based approaches utilize similarity information in a nearest-neighbor setting to classify time series data. Although approaches from this category provide accurate results, their performance degrades with long and noisy time series. On the other hand, feature-based approaches extract features to deal with the limitations of instance-based approaches; however, these approaches work with predefined features and may not be successful in certain classification problems. This study proposes a time series classification approach that benefits from both scale-space theory and bag-of-features technique. The method starts with finding the scale-space extrema points (i.e. key points) of each time series according to the SiZer (SIgnificant ZERo crossings of the derivatives) method, and then proceeds to create the local features set around these points. After extraction of the local features from each key point, a bag-of-features representation for each time series is constructed as the summary of the key point characteristics. We evaluate the success of the proposed representation on time series classification problems from various domains. Our experimental results show that our proposal provides competitive results compared to widely used approaches in the literature.

}, keywords = {Bag-of-Features Technique, Feature-based Classification, Scale-Space Theory, SiZer, time series classification}, issn = {2215-0986}, doi = {https://doi.org/10.1016/j.jestch.2021.03.017}, url = {https://www.sciencedirect.com/science/article/pii/S2215098621000823}, author = {Tayip Altay and Mustafa Gokce Baydogan} } @article {1421, title = {Randomized trees for time series representation and similarity}, journal = {Pattern Recognition}, volume = {120}, year = {2021}, pages = {108097}, abstract = {

Most of the temporal data mining tasks require representations to capture important characteristics of time series. Representation learning is challenging when time series differ in distributional characteristics and/or show irregularities such as varying lengths and missing observations. Moreover, when time series are multivariate, interactions between variables should be modeled efficiently. This study proposes a unified, flexible time series representation learning framework for both univariate and multivariate time series called Rand-TS. Rand-TS models density characteristics of each time series as a time-varying Gaussian distribution using random decision trees and embeds density information into a sparse vector. Rand-TS can work with time series of various lengths and missing observations, furthermore, it allows using customized features. We illustrate the classification performance of Rand-TS on 113 univariate, 19 multivariate along with 15 univariate time series with varying lengths from UCR database. The results show that in addition to its flexibility, Rand-TS provides competitive classification performance.

}, keywords = {Classification, Random trees, Representation learning, Time series}, issn = {0031-3203}, doi = {https://doi.org/10.1016/j.patcog.2021.108097}, url = {https://www.sciencedirect.com/science/article/pii/S0031320321002843}, author = {Berk G{\"o}rg{\"u}l{\"u} and Mustafa Gokce Baydogan} } @article {1419, title = {The benefits of target relations: A comparison of multitask extensions and classifier chains}, journal = {Pattern Recognition}, volume = {107}, year = {2020}, pages = {107507}, abstract = {

Multitask (multi-target or multi-output) learning (MTL) deals with simultaneous prediction of several outputs. MTL approaches rely on the optimization of a joint score function over the targets. However, defining a joint score in global models is problematic when the target scales are different. To address such problems, single target (i.e. local) learning strategies are commonly employed. Here we propose alternative tree-based learning strategies to handle the issue with target scaling in global models, and to identify the learning order for chaining operations in local models. In the first proposal, the problems with target scaling are resolved using alternative splitting strategies which consider the learning tasks in a multi-objective optimization framework. The second proposal deals with the problem of ordering in the chaining strategies. We introduce an alternative estimation strategy, minimum error chain policy, that gradually expands the input space using the estimations that approximate to true characteristics of outputs, namely out-of-bag estimations in tree-based ensemble framework. Our experiments on benchmark datasets illustrate the success of the proposed multitask extension of trees compared to the decision trees with de facto design especially for datasets with large number of targets. In line with that, minimum error chain policy improves the performance of the state-of-the-art chaining policies.

}, keywords = {Classifier chains, Ensemble learning, Multi-objective trees, Multitask learning, Stacking}, issn = {0031-3203}, doi = {https://doi.org/10.1016/j.patcog.2020.107507}, url = {https://www.sciencedirect.com/science/article/pii/S0031320320303101}, author = {Esra Ad{\i}yeke and Mustafa Gokce Baydogan} } @article {1415, title = {Classification of generic system dynamics model outputs via supervised time series pattern discovery}, journal = {Turkish Journal of Electrical Engineering \& Computer Sciences}, volume = {27}, year = {2019}, pages = {832{\textendash}846}, author = {Edali, Mert and Mustafa Gokce Baydogan and G{\"o}nen{\c c} Y{\"u}cel} } @article {1418, title = {Estimation of Influence Distance of Bus Stops Using Bus GPS Data and Bus Stop Properties}, journal = {IEEE Transactions on Intelligent Transportation Systems}, volume = {20}, year = {2019}, pages = {4635-4642}, doi = {10.1109/TITS.2019.2909645}, author = {Gokasar, Ilgin and Cetinel, Yigit and Mustafa Gokce Baydogan} } @article {1157, title = {Autoregressive forests for multivariate time series modeling}, journal = {Pattern Recognition}, volume = {73}, year = {2018}, pages = {202 - 215}, keywords = {Classification, Ensemble learning, Multivariate time series, Time series representation, Vector autoregression}, issn = {0031-3203}, doi = {https://doi.org/10.1016/j.patcog.2017.08.016}, url = {http://www.sciencedirect.com/science/article/pii/S0031320317303291}, author = {Kerem Sinan Tuncel and Mustafa Gokce Baydogan} } @inbook {1187, title = {Affect measurement: A roadmap through approaches, technologies, and data analysis}, booktitle = {Emotions and Affect in Human Factors and Human-Computer Interaction}, year = {2017}, pages = {255{\textendash}288}, publisher = {Elsevier}, organization = {Elsevier}, author = {Gonzalez-Sanchez, Javier and Mustafa Gokce Baydogan and Chavez-Echeagaray, Maria Elena and Atkinson, Robert K and Burleson, Winslow} } @conference {1167, title = {A scale-space theory and bag-of-features based time series classification method}, booktitle = {Signal Processing and Communications Applications Conference (SIU), 2017 25th}, year = {2017}, publisher = {IEEE}, organization = {IEEE}, author = {Altay, Tayip and Mustafa Gokce Baydogan} } @article {1063, title = {Time series representation and similarity based on local autopatterns}, journal = {Data Mining and Knowledge Discovery}, volume = {30}, year = {2016}, month = {03/2016}, pages = {476{\textendash}509}, abstract = {

Time series data mining has received much greater interest along with the increase in temporal data sets from different domains such as medicine, finance, multimedia, etc. Representations are important to reduce dimensionality and generate useful similarity measures. High-level representations such as Fourier transforms, wavelets, piecewise polynomial models, etc., were considered previously. Recently, autoregressive kernels were introduced to reflect the similarity of the time series. We introduce a novel approach to model the dependency structure in time series that generalizes the concept of autoregression to local autopatterns. Our approach generates a pattern-based representation along with a similarity measure called learned pattern similarity (LPS). A tree-based ensemble-learning strategy that is fast and insensitive to parameter settings is the basis for the approach. Then, a robust similarity measure based on the learned patterns is presented. This unsupervised approach to represent and measure the similarity between time series generally applies to a number of data mining tasks (e.g., clustering, anomaly detection, classification). Furthermore, an embedded learning of the representation avoids pre-defined features and an extraction step which is common in some feature-based approaches. The method generalizes in a straightforward manner to multivariate time series. The effectiveness of LPS is evaluated on time series classification problems from various domains. We compare LPS to eleven well-known similarity measures. Our experimental results show that LPS provides fast and competitive results on benchmark datasets from several domains. Furthermore, LPS provides a research direction and template approach that breaks from the linear dependency models to potentially foster other promising nonlinear approaches.

}, issn = {1573-756X}, author = {Mustafa Gokce Baydogan and Runger, George} } @article {657, title = {Learning a symbolic representation for multivariate time series classification}, journal = {Data Mining and Knowledge Discovery}, volume = {29}, year = {2015}, month = {03/2015}, pages = {400-422}, publisher = {Springer US}, abstract = {

Multivariate time series (MTS) classification has gained importance with the increase in the number of temporal datasets in different domains (such as medicine, finance, multimedia, etc.). Similarity-based approaches, such as nearest-neighbor classifiers, are often used for univariate time series, but MTS are characterized not only by individual attributes, but also by their relationships. Here we provide a classifier based on a new symbolic representation for MTS (denoted as SMTS) with several important elements. SMTS considers all attributes of MTS simultaneously, rather than separately, to extract information contained in the relationships. Symbols are learned from a supervised algorithm that does not require pre-defined intervals, nor features. An elementary representation is used that consists of the time index, and the values (and first differences for numerical attributes) of the individual time series as columns. That is, there is essentially no feature extraction (aside from first differences) and the local series values are fused to time position through the time index. The initial representation of raw data is quite simple conceptually and operationally. Still, a tree-based ensemble can detect interactions in the space of the time index and time values and this is exploited to generate a high-dimensional codebook from the terminal nodes of the trees. Because the time index is included as an attribute, each MTS is learned to be segmented by time, or by the value of one of its attributes. The codebook is processed with a second ensemble where now implicit feature selection is exploited to handle the high-dimensional input. The constituent properties produce a distinctly different algorithm. Moreover, MTS with nominal and missing values are handled efficiently with tree learners. Experiments demonstrate the effectiveness of the proposed approach in terms of accuracy and computation times in a large collection multivariate (and univariate) datasets.

}, keywords = {codebook, Decision trees, supervised learning}, issn = {1384-5810}, doi = {10.1007/s10618-014-0349-y}, url = {http://dx.doi.org/10.1007/s10618-014-0349-y}, author = {Mustafa Gokce Baydogan and Runger, George} } @article {908, title = {Metabolic connectivity as index of verbal working memory}, journal = {Journal of cerebral blood flow and metabolism}, volume = {35}, year = {2015}, month = {July}, pages = {1122{\textemdash}1126}, issn = {0271-678X}, doi = {10.1038/jcbfm.2015.40}, url = {http://dx.doi.org/10.1038/jcbfm.2015.40}, author = {Zou, Na and Chetelat, Gael and Mustafa Gokce Baydogan and Li, Jing and Fischer, Florian U and Titov, Dmitry and Dukart, Juergen and Fellgiebel, Andreas and Schreckenberger, Mathias and Yakushev, Igor} } @article {941, title = {A Transfer Learning Approach for Predictive Modeling of Degenerate Biological Systems}, journal = {Technometrics}, volume = {57}, year = {2015}, pages = {362-373}, doi = {10.1080/00401706.2015.1044117}, url = {http://dx.doi.org/10.1080/00401706.2015.1044117}, author = {Na Zou and Yun Zhu and Ji Zhu and Mustafa Gokce Baydogan and Wei Wang and Jing Li} } @article {35, title = {SMT: Sparse multivariate tree}, journal = {Statistical Analysis and Data Mining}, volume = {7}, year = {2014}, month = {02/2014}, pages = {53-69}, abstract = {

A multivariate decision tree attempts to improve upon the single variable split in a traditional tree. With the increase in datasets with many features and a small number of labeled instances in a variety of domains (bioinformatics, text mining, etc.), a traditional tree-based approach with a greedy variable selection at a node may omit important information. Therefore, the recursive partitioning idea of a simple decision tree combined with the intrinsic feature selection of L1 regularized logistic regression (LR) at each node is a natural choice for a multivariate tree model that is simple, but broadly applicable. This natural solution leads to the sparse multivariate tree (SMT) considered here. SMT can naturally handle non-time-series data and is extended to handle time-series classification problems with the power of extracting interpretable temporal patterns (e.g., means, slopes, and deviations). Binary L1 regularized LR models are used here for binary classification problems. However, SMT may be extended to solve multiclass problems with multinomial LR models. The accuracy and computational efficiency of SMT is compared to a large number of competitors on time series and non-time-series data.

}, keywords = {decision tree, feature extraction, fused Lasso, Lasso, time series classification}, issn = {1932-1872}, doi = {10.1002/sam.11208}, url = {http://dx.doi.org/10.1002/sam.11208}, author = {Houtao Deng and Mustafa Gokce Baydogan and George Runger} } @conference {39, title = {Affect Recognition in Learning Scenarios: Matching Facial-and BCI-Based Values}, booktitle = {13th IEEE International Conference on Advanced Learning Technologies (ICALT{\textquoteright}13)}, year = {2013}, publisher = {IEEE}, organization = {IEEE}, author = {Javier Gonzalez-Sanchez and Maria Elena Chavez-Echeagaray and Lijia Lin and Mustafa Gokce Baydogan and Robert Christopherson and David Gibson and Robert Atkinson and Winslow Burleson} } @article {33, title = {A Bag-of-Features Framework to Classify Time Series}, journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence}, volume = {35}, year = {2013}, pages = {2796-2802}, abstract = {

Time series classification is an important task with many challenging applications. A nearest neighbor (NN) classifier with dynamic time warping (DTW) distance is a strong solution in this context. On the other hand, feature-based approaches have been proposed as both classifiers and to provide insight into the series, but these approaches have problems handling translations and dilations in local patterns. Considering these shortcomings, we present a framework to classify time series based on a bag-of-features representation (TSBF). Multiple subsequences selected from random locations and of random lengths are partitioned into shorter intervals to capture the local information. Consequently, features computed from these subsequences measure properties at different locations and dilations when viewed from the original series. This provides a feature-based approach that can handle warping (although differently from DTW). Moreover, a supervised learner (that handles mixed data types, different units, etc.) integrates location information into a compact codebook through class probability estimates. Additionally, relevant global features can easily supplement the codebook. TSBF is compared to NN classifiers and other alternatives (bag-of-words strategies, sparse spatial sample kernels, shapelets). Our experimental results show that TSBF provides better results than competitive methods on benchmark datasets from the UCR time series database.

}, keywords = {codebook, feature extraction, supervised learning}, issn = {0162-8828}, doi = {http://doi.ieeecomputersociety.org/10.1109/TPAMI.2013.72}, author = {Mustafa Gokce Baydogan and George Runger and Eugene Tuv} } @article {34, title = {Toward Development of Adaptive Service-Based Software Systems}, journal = {IEEE Transactions on Services Computing}, volume = {2}, year = {2009}, pages = {247-260}, abstract = {

The rapid adoption of service-oriented architecture (SOA) in many large-scale distributed applications requires the development of adaptive service-based software systems (ASBS) with the capability of monitoring the changing system status, analyzing, and controlling tradeoffs among various quality-of-service (QoS) aspects, and adapting service configurations to satisfy multiple QoS requirements simultaneously. In this paper, our results toward the development of adaptive service-based software systems are presented. The formulation of activity-state-QoS (ASQ) models and how to use the data from controlled experiments to establish ASQ models for capturing the cause-effect dynamics among service activities, system resource states, and QoS in service-based systems are presented. Then, QoS monitoring modules based on ASQ models and SOA-compliant simulation models are developed to support the validation of the ASBS design. The main idea for developing QoS adaptation modules based on ASQ models is discussed. An experiment based on a voice communication service is used to illustrate our results.

}, keywords = {Design concepts, distributed/Internet-based software engineering tools and techniques, methodologies, modeling methodologies, quality of services, services systems}, issn = {1939-1374}, doi = {http://doi.ieeecomputersociety.org/10.1109/TSC.2009.17}, author = {Stephen S. Yau and Nong Ye and Hessam S. Sarjoughian and Dazhi Huang and Auttawut Roontiva and Mustafa Gokce Baydogan and Mohammed A. Muqsith} }