@inproceedings{2479, author = {{Mohr, Felix and Wever, Marcel Dominik and Hüllermeier, Eyke and Faez, Amin}}, booktitle = {{SCC}}, location = {{San Francisco, CA, USA}}, publisher = {{IEEE}}, title = {{{(WIP) Towards the Automated Composition of Machine Learning Services}}}, doi = {{10.1109/SCC.2018.00039}}, year = {{2018}}, } @unpublished{19524, abstract = {{Object ranking is an important problem in the realm of preference learning. On the basis of training data in the form of a set of rankings of objects, which are typically represented as feature vectors, the goal is to learn a ranking function that predicts a linear order of any new set of objects. Current approaches commonly focus on ranking by scoring, i.e., on learning an underlying latent utility function that seeks to capture the inherent utility of each object. These approaches, however, are not able to take possible effects of context-dependence into account, where context-dependence means that the utility or usefulness of an object may also depend on what other objects are available as alternatives. In this paper, we formalize the problem of context-dependent ranking and present two general approaches based on two natural representations of context-dependent ranking functions. Both approaches are instantiated by means of appropriate neural network architectures, which are evaluated on suitable benchmark task.}}, author = {{Pfannschmidt, Karlson and Gupta, Pritha and Hüllermeier, Eyke}}, booktitle = {{arXiv:1803.05796}}, title = {{{Deep Architectures for Learning Context-dependent Ranking Functions}}}, year = {{2018}}, } @inproceedings{2857, author = {{Mohr, Felix and Lettmann, Theodor and Hüllermeier, Eyke and Wever, Marcel Dominik}}, booktitle = {{Proceedings of the 1st ICAPS Workshop on Hierarchical Planning}}, location = {{Delft, Netherlands}}, pages = {{31--39}}, publisher = {{AAAI}}, title = {{{Programmatic Task Network Planning}}}, year = {{2018}}, } @article{24150, author = {{Ramaswamy, Arunselvan and Bhatnagar, Shalabh}}, journal = {{IEEE Transactions on Automatic Control}}, number = {{6}}, pages = {{2614--2620}}, publisher = {{IEEE}}, title = {{{Stability of stochastic approximations with “controlled markov” noise and temporal difference learning}}}, volume = {{64}}, year = {{2018}}, } @article{24151, author = {{Demirel, Burak and Ramaswamy, Arunselvan and Quevedo, Daniel E and Karl, Holger}}, journal = {{IEEE Control Systems Letters}}, number = {{4}}, pages = {{737--742}}, publisher = {{IEEE}}, title = {{{Deepcas: A deep reinforcement learning algorithm for control-aware scheduling}}}, volume = {{2}}, year = {{2018}}, } @inproceedings{2471, author = {{Mohr, Felix and Wever, Marcel Dominik and Hüllermeier, Eyke}}, booktitle = {{SCC}}, location = {{San Francisco, CA, USA}}, publisher = {{IEEE Computer Society}}, title = {{{On-The-Fly Service Construction with Prototypes}}}, doi = {{10.1109/SCC.2018.00036}}, year = {{2018}}, } @article{3402, abstract = {{In machine learning, so-called nested dichotomies are utilized as a reduction technique, i.e., to decompose a multi-class classification problem into a set of binary problems, which are solved using a simple binary classifier as a base learner. The performance of the (multi-class) classifier thus produced strongly depends on the structure of the decomposition. In this paper, we conduct an empirical study, in which we compare existing heuristics for selecting a suitable structure in the form of a nested dichotomy. Moreover, we propose two additional heuristics as natural completions. One of them is the Best-of-K heuristic, which picks the (presumably) best among K randomly generated nested dichotomies. Surprisingly, and in spite of its simplicity, it turns out to outperform the state of the art.}}, author = {{Melnikov, Vitalik and Hüllermeier, Eyke}}, issn = {{1573-0565}}, journal = {{Machine Learning}}, title = {{{On the effectiveness of heuristics for learning nested dichotomies: an empirical analysis}}}, doi = {{10.1007/s10994-018-5733-1}}, year = {{2018}}, } @article{3510, abstract = {{Automated machine learning (AutoML) seeks to automatically select, compose, and parametrize machine learning algorithms, so as to achieve optimal performance on a given task (dataset). Although current approaches to AutoML have already produced impressive results, the field is still far from mature, and new techniques are still being developed. In this paper, we present ML-Plan, a new approach to AutoML based on hierarchical planning. To highlight the potential of this approach, we compare ML-Plan to the state-of-the-art frameworks Auto-WEKA, auto-sklearn, and TPOT. In an extensive series of experiments, we show that ML-Plan is highly competitive and often outperforms existing approaches.}}, author = {{Mohr, Felix and Wever, Marcel Dominik and Hüllermeier, Eyke}}, issn = {{1573-0565}}, journal = {{Machine Learning}}, keywords = {{AutoML, Hierarchical Planning, HTN planning, ML-Plan}}, location = {{Dublin, Ireland}}, pages = {{1495--1515}}, publisher = {{Springer}}, title = {{{ML-Plan: Automated Machine Learning via Hierarchical Planning}}}, doi = {{10.1007/s10994-018-5735-z}}, year = {{2018}}, } @inproceedings{3552, author = {{Mohr, Felix and Wever, Marcel Dominik and Hüllermeier, Eyke}}, booktitle = {{Proceedings of the Symposium on Intelligent Data Analysis}}, location = {{‘s-Hertogenbosch, the Netherlands}}, title = {{{Reduction Stumps for Multi-Class Classification}}}, doi = {{10.1007/978-3-030-01768-2_19}}, year = {{2018}}, } @inproceedings{3852, abstract = {{In automated machine learning (AutoML), the process of engineering machine learning applications with respect to a specific problem is (partially) automated. Various AutoML tools have already been introduced to provide out-of-the-box machine learning functionality. More specifically, by selecting machine learning algorithms and optimizing their hyperparameters, these tools produce a machine learning pipeline tailored to the problem at hand. Except for TPOT, all of these tools restrict the maximum number of processing steps of such a pipeline. However, as TPOT follows an evolutionary approach, it suffers from performance issues when dealing with larger datasets. In this paper, we present an alternative approach leveraging a hierarchical planning to configure machine learning pipelines that are unlimited in length. We evaluate our approach and find its performance to be competitive with other AutoML tools, including TPOT.}}, author = {{Wever, Marcel Dominik and Mohr, Felix and Hüllermeier, Eyke}}, booktitle = {{ICML 2018 AutoML Workshop}}, keywords = {{automated machine learning, complex pipelines, hierarchical planning}}, location = {{Stockholm, Sweden}}, title = {{{ML-Plan for Unlimited-Length Machine Learning Pipelines}}}, year = {{2018}}, } @inproceedings{2109, abstract = {{In multinomial classification, reduction techniques are commonly used to decompose the original learning problem into several simpler problems. For example, by recursively bisecting the original set of classes, so-called nested dichotomies define a set of binary classification problems that are organized in the structure of a binary tree. In contrast to the existing one-shot heuristics for constructing nested dichotomies and motivated by recent work on algorithm configuration, we propose a genetic algorithm for optimizing the structure of such dichotomies. A key component of this approach is the proposed genetic representation that facilitates the application of standard genetic operators, while still supporting the exchange of partial solutions under recombination. We evaluate the approach in an extensive experimental study, showing that it yields classifiers with superior generalization performance.}}, author = {{Wever, Marcel Dominik and Mohr, Felix and Hüllermeier, Eyke}}, booktitle = {{Proceedings of the Genetic and Evolutionary Computation Conference, GECCO 2018, Kyoto, Japan, July 15-19, 2018}}, keywords = {{Classification, Hierarchical Decomposition, Indirect Encoding}}, location = {{Kyoto, Japan}}, publisher = {{ACM}}, title = {{{Ensembles of Evolved Nested Dichotomies for Classification}}}, doi = {{10.1145/3205455.3205562}}, year = {{2018}}, } @unpublished{17713, author = {{Wever, Marcel Dominik and Mohr, Felix and Hüllermeier, Eyke}}, publisher = {{Arxiv}}, title = {{{Automated Multi-Label Classification based on ML-Plan}}}, year = {{2018}}, } @unpublished{17714, author = {{Mohr, Felix and Wever, Marcel Dominik and Hüllermeier, Eyke}}, title = {{{Automated machine learning service composition}}}, year = {{2018}}, } @misc{5693, author = {{Graf, Helena}}, publisher = {{Universität Paderborn}}, title = {{{Ranking of Classification Algorithms in AutoML}}}, year = {{2018}}, } @misc{5936, author = {{Scheibl, Manuel}}, publisher = {{Universität Paderborn}}, title = {{{Learning about learning curves from dataset properties}}}, year = {{2018}}, } @inbook{6423, author = {{Schäfer, Dirk and Hüllermeier, Eyke}}, booktitle = {{Discovery Science}}, isbn = {{9783030017705}}, issn = {{0302-9743}}, pages = {{161--175}}, publisher = {{Springer International Publishing}}, title = {{{Preference-Based Reinforcement Learning Using Dyad Ranking}}}, doi = {{10.1007/978-3-030-01771-2_11}}, year = {{2018}}, } @proceedings{10591, editor = {{Abiteboul, S. and Arenas, M. and Barceló, P. and Bienvenu, M. and Calvanese, D. and David, C. and Hull, R. and Hüllermeier, Eyke and Kimelfeld, B. and Libkin, L. and Martens, W. and Milo, T. and Murlak, F. and Neven, F. and Ortiz, M. and Schwentick, T. and Stoyanovich, J. and Su, J. and Suciu, D. and Vianu, V. and Yi, K.}}, number = {{1}}, pages = {{1--29}}, title = {{{Research Directions for Principles of Data Management}}}, volume = {{7}}, year = {{2018}}, } @inbook{10783, author = {{Couso, Ines and Hüllermeier, Eyke}}, booktitle = {{Frontiers in Computational Intelligence}}, editor = {{Mostaghim, Sanaz and Nürnberger, Andreas and Borgelt, Christian}}, pages = {{31--46}}, publisher = {{Springer}}, title = {{{Statistical Inference for Incomplete Ranking Data: A Comparison of two likelihood-based estimators}}}, year = {{2018}}, } @article{16038, author = {{Schäfer, D. and Hüllermeier, Eyke}}, journal = {{Machine Learning}}, number = {{5}}, pages = {{903--941}}, title = {{{Dyad ranking using Plackett-Luce models based on joint feature representations}}}, volume = {{107}}, year = {{2018}}, } @inproceedings{10145, author = {{Ahmadi Fahandar, Mohsen and Hüllermeier, Eyke}}, booktitle = {{Proc. 32 nd AAAI Conference on Artificial Intelligence (AAAI)}}, pages = {{2951--2958}}, title = {{{Learning to Rank Based on Analogical Reasoning}}}, year = {{2018}}, }