Jerzy Swiatek

GPU-Based Massive Parallel Kawasaki Kinetics in the Dynamic Monte Carlo Simulations of Lipid Nanodomains

Multicomponent lipid membranes in the liquid phase exhibit dynamic lateral heterogeneities which play an important role in specific cell membrane functions. A GPU-based parallel algorithm for two-dimensional lattice Dynamic Monte Carlo simulations of nanodomain formation in binary lipid membranes was developed and tested. Speedups of up to 50-times over CPU-based calculations were achieved, and simulations employing lattices of up to 1800 × 1800 sites were performed. The existence of large nonregular lipid domains of sizes up to 160 nm was demonstrated. This reveals the necessity to employ lattices of at least ∼900 × 900 sites to study the lateral lipid organization in complex lipid membranes.

Global Identification of Complex Systems with Cascade Structure

In this paper, a problem of input-output complex systems identification is presented. The description of a complex system is given by a description of each system element and structure. Local and global identification problems are formulated, and a different approach to the identification problem is discussed. Based on the multi-criteria concept, the globally optimal model with respect to the quality of local models is presented.

Empirical Mode Decomposition Based Denoising Algorithm for Fibre Optical Gyroscope Measurement

The paper presents a new method to signal denoising based on Empirical Mode Decomposition and sparse optimization with application to fiber optical gyroscope measurement. The conventional approaches to signal denoising designed for Empirical Mode Decomposition are partial reconstruction and thresholding. Inspired by the second one, we propose a novel method that extends the performance the conventional methods. Our method based on the concept of sparse optimization. To validate the proposed approach, we test its performance for the real signal acquired from fiber optical gyroscope.

Ensemble SVM for imbalanced data and missing values in postoperative risk management

In this work, we propose the ensemble SVM that solves the problem of missing values of attributes and the imbalanced data phenomenon in the domain of postoperative risk management. Contrary to the other approaches the our solution effectively deals with the problems of high percentage of unknown values of the features. The problem of imbalanced data is solved by applying the cost-sensitive SVM as a base classifier of an ensemble, The quality of the proposed classifier is examined on a real-life dataset.