Search Results

You are looking at 1 - 10 of 101 items for

  • Keyword: Time series x
Clear All Modify Search
Open access

Lukasz Malinski

Abstract

The paper presents new approach to estimation of the coefficients of an elementary bilinear time series model (EB). Until now, a lot of authors have considered different identifiability conditions for EB models which implicated different identifiability ranges for the model coefficient. However, all of these ranges have a common feature namely they are significantly narrower than the stability range of the EB model. This paper proposes a simple but efficient solution which makes an estimation of the EB model coefficient possible within its entire stability range.

Open access

Tomasz Niedzielski

Empirical Hydrologic Predictions for Southwestern Poland and Their Relation to Enso Teleconnections

Recent investigations confirm meaningful but weak teleconnections between the El Niño/Southern Oscillation (ENSO) and hydrology in some European regions. In particular, this finding holds for Polish riverflows in winter and early spring as inferred from integrating numerous geodetic, geophysical and hydrologic time series. The purpose of this study is to examine whether such remote teleconnections may have an influence on hydrologic forecasting. The daily discharge time series from southwestern (SW) Poland spanning the time interval from 1971 to 2006 are examined. A few winter and spring peak flows are considered and the issue of their predictability using empirical forecasting is addressed. Following satisfactory prediction performance reported elsewhere, the multivariate autoregressive method is used and its modification based on the finite impulse response filtering is proposed. The initial phases of peak flows are rather acceptably forecasted but the accuracy of predictions in the vicinity of local maxima of the hydrographs is poorer. It has been hypothesized that ENSO signal slightly influences the predictability of winter and early spring floods in SW Poland. The predictions of flood wave maxima are the most accurate for floods preceded by normal states, less accurate for peak flows after La Niño episodes and highly inaccurate for peak flows preceded by El Niño events. Such a finding can be interpreted in terms of intermittency. Before peak flows preceded by El Niño there are temporarily persistent low flows followed by a consecutive melting leading to a considerable intermittency and hence to difficulties in forecasting. Before peak flows preceded by La Niño episodes there exist ENSO-related positive temperature and precipitation anomalies in SW Poland causing lower, but still considerable, intermittency and thus better, but not entirely correct, predictability of hydrologic time series.

Open access

Ratnesh Gautam and Anand K. Sinha

Abstract

Evapotranspiration is the one of the major role playing element in water cycle. More accurate measurement and forecasting of Evapotranspiration would enable more efficient water resources management. This study, is therefore, particularly focused on evapotranspiration modelling and forecasting, since forecasting would provide better information for optimal water resources management. There are numerous techniques of evapotranspiration forecasting that include autoregressive (AR) and moving average (MA), autoregressive moving average (ARMA), autoregressive integrated moving average (ARIMA), Thomas Feiring, etc. Out of these models ARIMA model has been found to be more suitable for analysis and forecasting of hydrological events. Therefore, in this study ARIMA models have been used for forecasting of mean monthly reference crop evapotranspiration by stochastic analysis. The data series of 102 years i.e. 1224 months of Bokaro District were used for analysis and forecasting. Different order of ARIMA model was selected on the basis of autocorrelation function (ACF) and partial autocorrelation (PACF) of data series. Maximum likelihood method was used for determining the parameters of the models. To see the statistical parameter of model, best fitted model is ARIMA (0, 1, 4) (0, 1, 1)12.

Open access

Silviu Rei, Dan Chicea, Beriliu Ilie and Sorin Olaru

Abstract

When performing data acquisition for a Dynamic Light Scattering experiment, one of the most important aspect is the filtering and conditioning of the electrical signal. The signal is amplified first and then fed as input for the analog digital convertor. As a result a digital time series is obtained. The frequency spectrum is computed by the logical unit offering the basis for further Dynamic Light Scattering analysis methods. This paper presents a simple setup that can accomplish the signal conditioning and conversion to a digital time series.

Open access

Dominika Dąbrowska, Marek Sołtysiak and Jan Waligóra

Abstract

The Ustroń S.A. Health Resort (southern Poland) uses iodide-bromide mineral waters taken from Middle and Upper Devonian limestones and dolomites with a mineralisation range of 110-130 g/dm3 for curative purposes. Two boreholes - U-3 and U3-A drilled in the early 1970s were exploited. The aim of this paper is to estimate changes in mineral water quality of the Ustroń Health Resort by taking into consideration chloride content in the water from the U-3 borehole. The data has included the results of monthly analyses of chlorides from 2005 to 2015 during the tests carried out by the Mining Department of the Health Resort. The triple exponential smoothing (ETS) function and the Seasonal Autoregressive Integrated Moving Average (SARIMA) method of modelling time series were used for the calculations. The ability to properly forecast mineral water quality can result in a good status of the exploitation borehole and a limited number of failures in the exploitation system. Because of the good management of health resorts, it is possible to acquire more satisfied customers. The main goal of the article involves the real-time forecast accuracy, obtained results show that the proposed methods are effective for such situations. Presented methods made it possible to obtain a 24-month point and interval forecast. The results of these analyses indicate that the chloride content is forecast to be in the range of 72 to 83 g/l from 2015 to 2017. While comparing the two methods of analysis, a narrower range of forecast values and, therefore, greater accuracy were obtained for the ETS function. The good performance of the ETS model highlights its utility compared with complicated physically based numerical models.

Open access

Mirosław Janik and Peter Bossew

Abstract

It is well known that the temporal dynamic of indoor and outdoor radon concentrations show complex patterns, which are partly not easy to interpret. Clearly, for physical reasons, they must be related to possibly variable conditions of radon generation, migration and atmospheric dispersion and accumulation. The aim of this study was to analyse long-time series of simultaneously measured indoor and outdoor radon concentrations, together with environmental quantities, which may act as control variables of Rn. The study was performed in Chiba, Japan, using two ionization chambers for parallel indoor and outdoor radon concentrations measurements over 4 years. Meteorological and seismic data were obtained from the Japan Metrological Agency (JMA).

Open access

Umi Mahmudah

Abstract

Nowadays it is getting harder for higher education graduates in finding a decent job. This study aims to predict the graduate unemployment in Indonesia by using autoregressive integrated moving average (ARIMA) model. A time series data of the graduate unemployment from 2005 to 2016 is analyzed. The results suggest that ARIMA (1,2,0) is the best model for forecasting analysis, where there is a tendency of increasing number for the next ten periods. Furthermore, the average of point forecast for the next 10 periods is about 1,266,179 while its minimum value is 1,012,861 the maximum values is 1,523,156. Overall, ARIMA (1,2,0) provides an adequate forecasting model so that there is no potential for improvement.

Open access

Dariusz Grzesica

Abstract

Decomposition of time series is the estimate and extraction of deterministic part of the series - trend, cyclical and seasonal fluctuations in the hope that the rest of the data, that is, theoretically, a random variable will be stationary random process. During the process of predicting the time series elements affects significantly on the determination of the future values, which are characterized by a low forecast error. Therefore, the purpose of this article is to identify the elements of the time series decomposition and to determine the extent to which they affect the forecasting process. Problems that often appear when you run the forecast and methods of building models and forecasts based on time series will be presented. Observations will be described on the basis of nonparametric time series modeling.

Open access

Janusz Bogusz, Anna Klos, Marta Gruszczynska and Maciej Gruszczynski

Abstract

In the modern geodesy the role of the permanent station is growing constantly. The proper treatment of the time series from such station lead to the determination of the reliable velocities. In this paper we focused on some pre-analysis as well as analysis issues, which have to be performed upon the time series of the North, East and Up components and showed the best, in our opinion, methods of determination of periodicities (by means of Singular Spectrum Analysis) and spatio-temporal correlations (Principal Component Analysis), that still exist in the time series despite modelling. Finally, the velocities of the selected European permanent stations with the associated errors determined following power-law assumption in the stochastic part is presented.

Open access

Pierre-Francois Marteau

Abstract

In the light of regularized dynamic time warping kernels, this paper re-considers the concept of a time elastic centroid for a set of time series. We derive a new algorithm based on a probabilistic interpretation of kernel alignment matrices. This algorithm expresses the averaging process in terms of stochastic alignment automata. It uses an iterative agglomerative heuristic method for averaging the aligned samples, while also averaging the times of their occurrence. By comparing classification accuracies for 45 heterogeneous time series data sets obtained by first nearest centroid/medoid classifiers, we show that (i) centroid-based approaches significantly outperform medoid-based ones, (ii) for the data sets considered, our algorithm, which combines averaging in the sample space and along the time axes, emerges as the most significantly robust model for time-elastic averaging with a promising noise reduction capability. We also demonstrate its benefit in an isolated gesture recognition experiment and its ability to significantly reduce the size of training instance sets. Finally, we highlight its denoising capability using demonstrative synthetic data. Specifically, we show that it is possible to retrieve, from few noisy instances, a signal whose components are scattered in a wide spectral band.