monitoring, predictive maintenance, and highly automated driving. However, obtaining labeled anomalous data can be a formidable
challenge, especially when anomalies exhibit temporal evolution. This paper introduces LATAM (Long short-term memory Autoencoder with Temporal Attention Mechanism) for few-shot anomaly detection, with the aim of enhancing detection performance in scenarios with limited labeled anomaly data. LATAM effectively captures temporal dependencies and emphasizes significant patterns in multivariate time series data. In our investigation, we
comprehensively evaluate LATAM against other anomaly detection models, particularly assessing its capability in few-shot learning
scenarios where we have minimal examples from the normal class and none from the anomalous class in the training data. Our
experimental results, derived from real-world photovoltaic inverter data, highlight LATAM’s superiority, showcasing a substantial
27% mean F1 score improvement, even when trained on a mere two-week dataset. Furthermore, LATAM demonstrates remarkable
results on the open-source SWaT dataset, achieving a 12% boost in accuracy with only two days of training data. Moreover, we
introduce a simple yet effective dynamic thresholding mechanism, further enhancing the anomaly detection capabilities of LATAM.
This underscores LATAM’s efficacy in addressing the challenges posed by limited labeled anomalies in practical scenarios and it
proves valuable for downstream tasks involving temporal representation and time series prediction, extending its utility beyond
anomaly detection applications. ER - TY - CONF AU - Mitchell, T. AU - Cohen, W. AU - Hruscha, E. AU - Talukdar, P. AU - Betteridge, J. AU - Carlson, A. AU - Dalvi, B. AU - Gardner, M. AU - Kisiel, B. AU - Krishnamurthy, J. AU - Lao, N. AU - Mazaitis, K. AU - Mohammad, T. AU - Nakashole, N. AU - Platanios, E. AU - Ritter, A. AU - Samadi, M. AU - Settles, B. AU - Wang, R. AU - Wijaya, D. AU - Gupta, A. AU - Chen, X. AU - Saparov, A. AU - Greaves, M. AU - Welling, J. A2 - T1 - Never-Ending Learning T2 - AAAI PB - CY - PY - 2015/ M2 - VL - IS - SP - EP - UR - http://www.cs.cmu.edu/~wcohen/pubs.html M3 - KW - learning KW - nell KW - ontology KW - semantic KW - toread L1 - SN - N1 - Papers by William W. Cohen N1 - AB - ER - TY - JOUR AU - Mnih, Volodymyr AU - Kavukcuoglu, Koray AU - Silver, David AU - Rusu, Andrei A. AU - Veness, Joel AU - Bellemare, Marc G. AU - Graves, Alex AU - Riedmiller, Martin AU - Fidjeland, Andreas K. AU - Ostrovski, Georg AU - Petersen, Stig AU - Beattie, Charles AU - Sadik, Amir AU - Antonoglou, Ioannis AU - King, Helen AU - Kumaran, Dharshan AU - Wierstra, Daan AU - Legg, Shane AU - Hassabis, Demis T1 - Human-level control through deep reinforcement learning JO - Nature PY - 2015/02 VL - 518 IS - 7540 SP - 529 EP - 533 UR - http://dx.doi.org/10.1038/nature14236 M3 - KW - deep KW - learning KW - toread L1 - SN - N1 - Human-level control through deep reinforcement learning - nature14236.pdf N1 - AB - ER - TY - CONF AU - Ring, Markus AU - Otto, Florian AU - Becker, Martin AU - Niebler, Thomas AU - Landes, Dieter AU - Hotho, Andreas A2 - ECMLPKDD2015 T1 - ConDist: A Context-Driven Categorical Distance Measure T2 - PB - CY - PY - 2015/ M2 - VL - IS - SP - EP - UR - M3 - KW - 2015 KW - categorical KW - data KW - learning KW - measure KW - myown KW - similarity KW - unsupervised L1 - SN - N1 - N1 - AB - ER - TY - CONF AU - Krompass, Denis AU - Nickel, Maximilian AU - Tresp, Volker A2 - T1 - Large-scale factorization of type-constrained multi-relational data T2 - International Conference on Data Science and Advanced Analytics, DSAA 2014, Shanghai, China, October 30 - November 1, 2014 PB - IEEE CY - PY - 2014/ M2 - VL - IS - SP - 18 EP - 24 UR - http://dx.doi.org/10.1109/DSAA.2014.7058046 M3 - 10.1109/DSAA.2014.7058046 KW - graph KW - knowledge KW - learning KW - toread L1 - SN - 978-1-4799-6991-3 N1 - dblp: BibTeX record conf/dsaa/KrompassNT14 N1 - AB - ER - TY - CHAP AU - Lehmann, Jens AU - Voelker, Johanna A2 - Lehmann, Jens A2 - Voelker, Johanna T1 - An Introduction to Ontology Learning T2 - Perspectives on Ontology Learning PB - AKA / IOS Press CY - PY - 2014/ VL - IS - SP - ix EP - xvi UR - http://jens-lehmann.org/files/2014/pol_introduction.pdf M3 - KW - introduction KW - learning KW - ontology L1 - SN - N1 - N1 - AB - ER - TY - JOUR AU - Kluegl, Peter AU - Toepfer, Martin AU - Lemmerich, Florian AU - Hotho, Andreas AU - Puppe, Frank T1 - Exploiting Structural Consistencies with Stacked Conditional Random Fields JO - Mathematical Methodologies in Pattern Recognition and Machine Learning Springer Proceedings in Mathematics & Statistics PY - 2013/ VL - 30 IS - SP - 111 EP - 125 UR - M3 - KW - 2013 KW - ie KW - learning KW - myown KW - references L1 - SN - N1 - N1 - AB - Conditional Random Fields (CRF) are popular methods for labeling unstructured or textual data. Like many machine learning approaches, these undirected graphical models assume the instances to be independently distributed. However, in real-world applications data is grouped in a natural way, e.g., by its creation context. The instances in each group often share additional structural consistencies. This paper proposes a domain-independent method for exploiting these consistencies by combining two CRFs in a stacked learning framework. We apply rule learning collectively on the predictions of an initial CRF for one context to acquire descriptions of its specific properties. Then, we utilize these descriptions as dynamic and high quality features in an additional (stacked) CRF. The presented approach is evaluated with a real-world dataset for the segmentation of references and achieves a significant reduction of the labeling error. ER - TY - GEN AU - Yu, Hsiang-Fu AU - Jain, Prateek AU - Kar, Purushottam AU - Dhillon, Inderjit S. A2 - T1 - Large-scale Multi-label Learning with Missing Labels JO - PB - AD - PY - 2013/ VL - IS - SP - EP - UR - http://arxiv.org/abs/1307.5101 M3 - KW - classification KW - kallimachos KW - label KW - large KW - learning KW - multi L1 - N1 - Large-scale Multi-label Learning with Missing Labels N1 - AB - The multi-label classification problem has generated significant interest in
recent years. However, existing approaches do not adequately address two key
challenges: (a) the ability to tackle problems with a large number (say
millions) of labels, and (b) the ability to handle data with missing labels. In
this paper, we directly address both these problems by studying the multi-label
problem in a generic empirical risk minimization (ERM) framework. Our
framework, despite being simple, is surprisingly able to encompass several
recent label-compression based methods which can be derived as special cases of
our method. To optimize the ERM problem, we develop techniques that exploit the
structure of specific loss functions - such as the squared loss function - to
offer efficient algorithms. We further show that our learning framework admits
formal excess risk bounds even in the presence of missing labels. Our risk
bounds are tight and demonstrate better generalization performance for low-rank
promoting trace-norm regularization when compared to (rank insensitive)
Frobenius norm regularization. Finally, we present extensive empirical results
on a variety of benchmark datasets and show that our methods perform
significantly better than existing label compression based methods and can
scale up to very large datasets such as the Wikipedia dataset. ER - TY - CONF AU - Mirowski, Piotr AU - Ranzato, Marc'Aurelio AU - LeCun, Yann A2 - of the NIPS 2010 Workshop on Deep Learning, Proceedings T1 - Dynamic Auto-Encoders for Semantic Indexing T2 - PB - CY - PY - 2010/ M2 - VL - IS - SP - EP - UR - http://yann.lecun.com/exdb/publis/pdf/mirowski-nipsdl-10.pdf M3 - KW - deep KW - kallimachos KW - lda KW - learning KW - model KW - toread L1 - SN - N1 - Neuer Tab N1 - AB - ER - TY - JOUR AU - Breiman, Leo T1 - Random Forests JO - Machine Learning PY - 2001/ VL - 45 IS - 1 SP - 5 EP - 32 UR - http://dx.doi.org/10.1023/A%3A1010933404324 M3 - 10.1023/A:1010933404324 KW - classification KW - ensemble KW - forest KW - learning KW - random L1 - SN - N1 - Random Forests - Springer N1 - AB - Random forests are a combination of tree predictors such that each tree depends on the values of a random vector sampled independently and with the same distribution for all trees in the forest. The generalization error for forests converges a.s. to a limit as the number of trees in the forest becomes large. The generalization error of a forest of tree classifiers depends on the strength of the individual trees in the forest and the correlation between them. Using a random selection of features to split each node yields error rates that compare favorably to ER -