The problem of change detection in nonstatioriary time series using
linear regression models is addressed. It is assumed that the data can by accurately described by a linear regression model with piece-wise constant parameters. Due to the limitations of some classical approaches, based upon the innovation of one autoregressive (AR) model, most algorithms for the change detection presented make use of two AR models: one is a reference model, and the other one is a current model updated via a sliding block. Changes are detected when a suitable “distance” between these two models is high. Three “distance” measures are considered in the paper: cepstral distance, log-likelihood ratio (justified by GLR) and a distance involving the cross-entropy of the two conditional probabilities laws (divergence test). Other methods based on the quadratic forms of Gaussian random variables are also discussed in the paper. Finally, a change detection algorithm using three models and the evolution of Akaike Information Criterion is presented. All the presented algorithms constituted the object of evaluation by multiple simulation and háve been used to change detection in some nonstationary financial and economic time series.
Based on the analysis performed in the Kura geomorphological sub-region of the Caspian coastal zone, it was determined that the sea level rise of 2.43 m in 1976-1996 caused an average annual rate of erosion processes about 75 times higher than accumulation processes. In 1996-2019, the stabilization and the following drop of the sea level of 1.39 m led to an average annual rate of accumulation processes up to 14 times higher than the average annual rate of erosion processes. This pattern is observed along the entire coastline of those periods. Erosion processes in the initial stage (1976-1996) increased the indentation and length of the coastline by almost 2 times, and in the next stage (1996-2019) the intensification of accumulation processes resulted in the smoothing and reduction of the coastline by 35 %. Statistical analysis of coastlines was carried out by processing Landsat satellite images for 1976-2019; each coastline was divided into 14 segments according to geomorphological zones and their morphostructures in the study area. The development forecast for the next 10 and 20 years is provided., Jeyhun Yashar oglu Gasimov., and Obsahuje bibliografii
The paper presents a new approach for a machine vibration analysis and health monitoring combining blind source separation (BSS) and change detection in source signals. So, the problem is translated from the space of the measurements to the space of independent sources, where the reduced number of components simplifies the monitoring problem and where the change detection methods are applied for scalar signals. The approach has been tested in simulation and the assessment on a real machine is presented in the last part of the paper.
Táto štúdia má za cieľ predstaviť a vysvetliť fenomén nedávno objavený, a tým je slepota k zmene (change blindness); neschopnosť detekovať markantné zmeny. Cieľom je popísať prečo a za akých podmienok neschopnosť detekovať zmeny vzniká. Slepotu k zmene možno indukovať prostredníctvom viacerých designov (flicker, one-shot, „škvrny“, prerušenie počas filmu). Techniky flicker a one-shot sú popísané detailne, poukazujúc na úlohu kognitívnych a exekutívnych procesov v procese detekovania zmien. Snaha poukázať na indukovanie fenoménu slepoty k zmene jednak v laboratórnom, jednak v prirodzenom prostredí je demonštrovaná prostredníctvom príkladov z mnohých výskumných štúdií. Preto implikácie ponúkané výskumom slepoty k zmene by mohli pomôcť porozumeniu spracovávania vizuálnych informácií, a takisto prispieť k objasneniu permanentnej dilemy v psychológii: úlohe spracovávania informácií zdola-hore a zhora-dole.