Prof. Dr.-Ing.

Bauckhage, Christian

Filters

10
9 1
Reset filters

Settings
Now showing 1 - 10 of 10
  • Publication
    Switching Dynamical Systems with Deep Neural Networks
    ( 2021-05-05)
    Ojeda, César
    ;
    Georgiev, Bogdan
    ;
    Cvejoski, Kostadin
    ;
    Schücker, Jannis
    ;
    Bauckhage, Christian
    ;
    Sánchez, Ramsés J.
    The problem of uncovering different dynamical regimes is of pivotal importance in time series analysis. Switching dynamical systems provide a solution for modeling physical phenomena whose time series data exhibit different dynamical modes. In this work we propose a novel variational RNN model for switching dynamics allowing for both non-Markovian and nonlinear dynamical behavior between and within dynamic modes. Attention mechanisms are provided to inform the switching distribution. We evaluate our model on synthetic and empirical datasets of diverse nature and successfully uncover different dynamical regimes and predict the switching dynamics.
  • Publication
    Auto Encoding Explanatory Examples with Stochastic Paths
    ( 2021-05-05)
    Ojeda, César
    ;
    Sánchez, Ramsés J.
    ;
    Cvejoski, Kostadin
    ;
    Schücker, Jannis
    ;
    Bauckhage, Christian
    ;
    Georgiev, Bogdan
    In this paper we ask for the main factors that determine a classifiers decision making process and uncover such factors by studying latent codes produced by auto-encoding frameworks. To deliver an explanation of a classifiers behaviour, we propose a method that provides series of examples highlighting semantic differences between the classifiers decisions. These examples are generated through interpolations in latent space. We introduce and formalize the notion of a semantic stochastic path, as a suitable stochastic process defined in feature (data) space via latent code interpolations. We then introduce the concept of semantic Lagrangians as a way to incorporate the desired classifiers behaviour and find that the solution of the associated variational problem allows for highli ghting differences in the classifier decision. Very importantly, within our framework the classifier is used as a black-box, and only its evaluation is required.
  • Publication
    Learning Deep Generative Models for Queuing Systems
    ( 2021)
    Ojeda, César
    ;
    Cvejoski, Kostadin
    ;
    Georgiev, Bogdan
    ;
    Bauckhage, Christian
    ;
    Schücker, Jannis
    ;
    Sánchez, Ramsés J.
    Modern society is heavily dependent on large scale client-server systems with applications ranging from Internet and Communication Services to sophisticated logistics and deployment of goods. To maintain and improve such a system, a careful study of client and server dynamics is needed e.g. response/service times, aver-age number of clients at given times, etc. To this end, one traditionally relies, within the queuing theory formalism, on parametric analysis and explicit distribution forms. However, parametric forms limit the models expressiveness and could struggle on extensively large datasets. We propose a novel data-driven approach towards queuing systems: the Deep Generative Service Times. Our methodology delivers a flexible and scalable model for service and response times. We leverage the representation capabilities of Recurrent Marked Point Processes for the temporal dynamics of clients, as well as Wasserstein Generative Adversarial Network techniques, to learn deep generative models which are able to represent complex conditional service time distributions. We provide extensive experimental analysis on both empirical and synthetic datasets, showing the effectiveness of the proposed models.
  • Publication
    Recurrent Adversarial Service Times
    ( 2019)
    Ojeda, César
    ;
    Cvejosky, Kostadin
    ;
    Sánchez, Ramsés J.
    ;
    Schücker, Jannis
    ;
    Georgiev, Bogdan
    ;
    Bauckhage, Christian
    Service system dynamics occur at the interplay between customer behaviour and a service provider's response. This kind of dynamics can effectively be modeled within the framework of queuing theory where customers' arrivals are described by point process models. However, these approaches are limited by parametric assumptions as to, for example, inter-event time distributions. In this paper, we address these limitations and propose a novel, deep neural network solution to the queuing problem. Our solution combines a recurrent neural network that models the arrival process with a recurrent generative adversarial network which models the service time distribution. We evaluate our methodology on various empirical datasets ranging from internet services (Blockchain, GitHub, Stackoverflow) to mobility service systems (New York taxi cab).
  • Publication
    Ising models for binary clustering via adiabatic quantum computing
    ( 2018)
    Bauckhage, Christian
    ;
    Brito, Eduardo
    ;
    Cvejoski, Kostadin
    ;
    Ojeda, César
    ;
    Sifa, Rafet
    ;
    Wrobel, Stefan
    Existing adiabatic quantum computers are tailored towards minimizing the energies of Ising models. The quest for implementations of pattern recognition or machine learning algorithms on such devices can thus be seen as the quest for Ising model (re-)formulations of their objective functions. In this paper, we present Ising models for the tasks of binary clustering of numerical and relational data and discuss how to set up corresponding quantum registers and Hamiltonian operators. In simulation experiments, we numerically solve the respective Schrödinger equations and observe our approaches to yield convincing results.
  • Publication
    Inverse dynamical inheritance in stack exchange taxonomies
    ( 2017)
    Ojeda, César
    ;
    Cvejoski, Kostadin
    ;
    Sifa, Rafet
    ;
    Bauckhage, Christian
    Question Answering websites are popular repositories of expert knowledge and cover areas as diverse as linguistics, computer science, or mathematics. Knowledge is commonly organized via user defined tags which implicitly create population folksonomies. However, the interplay between latent knowledge structures and the answering behavior of users has not been fully explored yet. Here, we propose a model of a dynamical tagging process guided by taxonomies, devise a robust algorithm that allow us to uncover hidden topic hierarchies, apply our method to analyze several Stack Exchange websites. Our results show that the dynamics of the system strongly correlate with uncovered taxonomies.
  • Publication
    Third party effect: Community based spreading in complex networks
    ( 2017)
    Bauckhage, Christian
    ;
    Ojeda, César
    ;
    Sifa, Rafet
    ;
    Agarwal, Shubham
    A substantial amount of network research has been devoted to the study of spreading processes and community detection without considering the role of communities in the characteristics of spreading processes. Here, we generalize the SIR model of epidemics by introducing a matrix of community infecting rates to capture the heterogeneous nature of the spreading as defined by the natural characteristics of communities. We find that the spreading capabilities of one community towards another is influenced by the internal behavior of third party communities. Our results provide insights into systems with rich information structure and into populations with diverse immunology responses.
  • Publication
    Investigating and forecasting user activities in newsblogs: A study of seasonality, volatility and attention burst
    ( 2017)
    Bauckhage, Christian
    ;
    Ojeda, César
    ;
    Sifa, Rafet
    The study of collective attention is a major topic in the area of Web science as we are interested to know how a particular news topic or meme is gaining or losing popularity over time. Recent research focused on developing methods which quantify the success and popularity of topics and studyied their dynamics over time. Yet, the aggregate behavior of users across content creation platforms has been largely ignored even though the popularity of news items is also linked to the way users interact with the Web platforms. In this paper, we present a novel framework of research which studies the shift of attentions of population over newsblogs. We concentrate on the commenting behavior of users for news articles which serves as a proxy for attention to Web content. We make use of methods from signal processing and econometrics to uncover patterns in the behavior of users which then allow us to simulate and hence to forecast the behavior of a population once an attention shift occurs. Studying a dataset of over 200 blogs with 14 million news posts, we found periodic regularities in the commenting behavior. Namely, cycles of 7 days as well as 24 days of activity which may be related to known scales of meme lifetimes.
  • Publication
    Towards German Word Embeddings: A Use Case with Predictive Sentiment Analysis
    ( 2017)
    Brito, Eduardo
    ;
    Sifa, Rafet
    ;
    Cvejoski, Kostadin
    ;
    Ojeda, César
    ;
    Bauckhage, Christian
    Despite the research boom on words embeddings and their text mining applications from the last years, the vast majority of publications focus only on the English language. Furthermore, hyperparameter tuning is a rarely well documented process (specially for non English text) that is necessary to obtain high quality word representations. In this work, we present how different hyperparameter combinations impact the resulting German word vectors and how these word representations can be part of more complex models. In particular, we perform first an intrinsic evaluation of our German word embeddings, which are later used within a predictive sentiment analysis model. The latter does not only serve as an extrinsic evaluation of the German word embeddings but also shows the feasibility of predic ting preferences only from document embeddings.
  • Publication
    Predicting retention in sandbox games with tensor factorization-based representation learning
    ( 2016)
    Sifa, Rafet
    ;
    Srikanth, Sridev
    ;
    Drachen, Anders
    ;
    Ojeda, César
    ;
    Bauckhage, Christian
    Major commercial (AAA) games increasingly transit to a semi-persistent or persistent format in order to extend the value of the game to the player, and to add new sources of revenue beyond basic retail sales. Given this shift in the design of AAA titles, game analytics needs to address new types of problems, notably the problem of forecasting future player behavior. This is because player retention is a key factor in driving revenue in semi-persistent titles, for example via downloadable content. This paper introduces a model for predicting retention of players in AAA games and provides a tensor-based spatio-temporal model for analyzing player trajectories in 3D games. We show how knowledge as to trajectories can help with predicting player retention. Furthermore, we describe two new algorithms for three way DEDICOM including a fast gradient method and a semi-nonnegative constrained method. These approaches are validated against a detailed behavioral data set from the AAA open-world game Just Cause 2.