Wyniki wyszukiwania - BazTech

1

Multilinear Filtering Based on a Hierarchical Structure of Covariance Matrices

Szwabe A., Misiorek P, Ciesielczyk M.

Schedae Informaticae

|

2015

|

Vol. 24

103--112

EN

We propose a novel model of multilinear filtering based on a hierarchical structure of covariance matrices – each matrix being extracted from the input tensor in accordance to a specific set-theoretic model of data generalization, such as derivation of expectation values. The experimental analysis results presented in this paper confirm that the investigated approaches to tensor-based data representation and processing outperform the standard collaborative filtering approach in the ‘cold-start’ personalized recommendation scenario (of very sparse input data). Furthermore, it has been shown that the proposed method is superior to standard tensor-based frameworks such as N-way Random Indexing (NRI) and Higher-Order Singular Value Decomposition (HOSVD) in terms of both the AUROC measure and computation time.

2

Holistic Entropy Reduction for Collaborative Filtering

Szwabe A., Misiorek P, Janasiewicz T, Walkowiak P.

Foundations of Computing and Decision Sciences

|

2014

|

Vol. 39, No. 3

209--229

EN

We propose a collaborative filtering (CF) method that uses behavioral data provided as propositions having the RDF-compliant form of (user X, likes, item Y ) triples. The method involves the application of a novel self-configuration technique for the generation of vector-space representations optimized from the information-theoretic perspective. The method, referred to as Holistic Probabilistic Modus Ponendo Ponens (HPMPP), enables reasoning about the likelihood of unknown facts. The proposed vector-space graph representation model is based on the probabilistic apparatus of quantum Information Retrieval and on the compatibility of all operators representing subjects, predicates, objects and facts. The dual graph-vector representation of the available propositional data enables the entropy-reducing transformation and supports the compositionality of mutually compatible representations. As shown in the experiments presented in the paper, the compositionality of the vector-space representations allows an HPMPP-based recommendation system to identify which of the unknown facts having the triple form (user X, likes, item Y ) are the most likely to be true in a way that is both effective and, in contrast to methods proposed so far, fully automatic.

3

Collaborative filtering based on bi-relational data representation

Szwabe A., Misiorek P, Ciesielczyk M, Jedrzejek C

Foundations of Computing and Decision Sciences

|

2013

|

Vol. 38, No. 1

67--83

EN

Widely-referenced approaches to collaborative filtering (CF) are based on the use of an input matrix that represents each user profile as a vector in a space of items and each item as a vector in a space of users. When the behavioral input data have the form of (userX, likes, itemY) and (userX, dislikes, itemY) triples one has to propose a representation of the user feedback data that is more suitable for the use of propositional data than the ordinary user-item ratings matrix. We propose to use an element-fact matrix, in which columns represent RDF-like behavioral data triples and rows represent users, items, and relations. By following such a triple-based approach to the bi-relational behavioral data representation we are able to improve the quality of collaborative filtering. One of the key findings of the research presented in this paper is that the proposed bi-relational behavioral data representation, while combined with reflective matrix processing, significantly outperforms state-of-the-art collaborative filtering methods based on the use of a ‘standard’ user-item matrix.

4

Critical analysis of existing approaches to wireless multi-service network design

Szwabe A., Walkowiak P.

Studia z Automatyki i Informatyki

|

2009

|

T. 34

73-84

EN

The paper surveys existing frameworks for design of wireless networks capable to serve delay-sensitive streams in parallel with delay-insensitive TCP-based data traffic. The problem of optimal control for wireless networks serving only rate-elastic flows is widely investigated by using models of Network Utility Maximization (NUM). The paper presents a survey of recent efforts to incorporate delay analysis into the NUM-oriented research that may possibly result in a more general optimization of multi-service networks. The arising key open issues are pointed out and discussed.

PL

Artykuł zawiera przegląd istniejących modeli sieci bezprzewodowych obsługujących zarówno wrażliwy na opóźnienie ruch strumieniowy, jak i elastyczny ruch wymagający wysokiej przepustowości. Problem optymalnej kontroli ruchu w sieciach bezprzewodowych obsługujących tylko ruch elastyczny jest dobrze osadzony w literaturze w postaci wielu prac traktujących o modelu NUM (Network Utility Maximization). W artykule skupiono się na propozycjach integrujących dotychczasowe rozwiązania oparte na modelu NUM z analizą wpływu opóźnienia na użyteczność sieci, w celu objęcia modelem ruchu heterogenicznego. W pracy omówione zostały również prawdopodobne dalsze kierunki rozwoju metod optymalizacji wykorzystania zasobów sieci.

5

Kompensacja wpływu czasu martwego na stabilność układu sterującego szybkością strumienia

Szwabe A.

Studia z Automatyki i Informatyki

|

2007

|

T. 32

155-172

PL

Artykuł prezentuje oryginalne podejście do problemu redukcji oscylacji chwilowej szybkości strumienia audiowizualnego transmitowanego w sieci IP. Założono zgodność modelowanego systemu adaptacji strumieni z powszechnie stosowanym zestawem protokołów RTP-RTCP, zaproponowano włączenie do systemu sterującego zmodyfikowanego predyktora Smitha. Układ obejmujący źródło strumienia, najsilniej obciążony węzeł sieci oraz odbiornik został przedstawiony w postaci modelu nieliniowego układu z czasem dyskretnym. Zaproponowano aproksymację modelu układu nieliniowego do postaci zgodnej z kryteriami LTI oraz zastosowanie elementów teorii układów LTI do analizy dynamicznych własności układu. Zaproponowana metoda optymalizacyjna umożliwia estymację wartości parametrów układu zapewniających jego stabilność oraz skuteczną redukcję wahań wartości chwilowej dwukierunkowego opóźnienia transmisji.

EN

The paper presents an original control-theoretic approach to a problem of stream rate oscillation cancellation. The main objective was to develop a method that is applicable to an audiovisual streaming system operating in an IP network. A proposed model represents functions of a real stream adaptation system based on the widely used RTP/RTCP protocol set. The proposed oscillation cancellation system includes a proportional controller supported by a modified Smith predictor. A discrete-time, nonlinear, time-variant model of a complete system (including a stream source, a network and a stream receiver) has been simplified into an LTI-compliant form. The transformed model has been used in analysis of the system's dynamic characteristics. The criteria for the optimal system's configuration refer to representation of the transfer function's roots on the Z-transform plane. It has been demonstrated how the proposed optimization methods enable development of a rate control module which effectively cancels stream rate oscillations.