Wyniki wyszukiwania - BazTech

1

Surgical phase classification and operative skill assessment through spatial context aware CNNs and time-invariant feature extracting autoencoders

Pradeep Chakka Sai, Sinha Neelam

Biocybernetics and Biomedical Engineering

|

2023

|

Vol. 43, no. 4

700--724

EN

Automated surgical video analysis promises improved healthcare. We propose novel spatial context aware combined loss function for end-to-end Encoder-Decoder training for Surgical Phase Classification (SPC) on laparoscopic cholecystectomy (LC) videos. Proposed loss function leverages on fine-grained class activation maps obtained from fused multi-layer Layer-CAM for supervised learning of SPC, obtaining improved Layer-CAM explanations. Post classification, we introduce graph theory to incorporate known hierarchies of surgical phases. We report peak SPC accuracy of 96.16%, precision of 94.08% and recall of 90.02% on public dataset Cholec80, with 7 phases. Our proposed method utilizes just 73.5% of parameters as against existing state-of-the-art methodology, achieving improvement of 0.5% in accuracy, 1.76% in precision with comparable recall, with an order less standard deviation. We also propose DNN based surgical skill assessment methodology. This approach utilizes surgical phase prediction scores from the final fully-connected layer of spatial-context aware classifier to form multi-channel temporal signal of surgical phases. Time-invariant representation is obtained from this temporal signal through time- and frequency-domain analyses. Autoencoder based time-invariant features are utilized for reconstruction and identification of prominent peaks in dissimilarity curves. We devise a surgical skill measure (SSM) based on spatial-context aware temporal-prominence-of-peaks curve. SSM values are expected to be high when executed skillfully, aligning with expert assessed GOALS metric. We illustrate this trend on Cholec80 and m2cai16-tool datasets, in comparison with GOALS metric. Concurrence in the trend of SSM with respect to GOALS metric is obtained on these test videos, making it a promising step towards automated surgical skill assessment.

2

Research on Earth rotation and geodynamics in Poland in 2019–2022

Bogusz Janusz, Brzezinski Aleksander, Godah Walyeldeen, Nastula Jolanta

Advances in Geodesy and Geoinformation

|

2023

|

Vol. 72, no. 2

art. no. e41, 2023

EN

This paper summarizes the activity of the chosen Polish geodetic research teams in 2019–2022 in the fields of the Earth rotation and geodynamics. This publication has been prepared for the needs of the presentation of Polish scientists’ activities on the 28th International Union of Geodesy and Geodynamics General Assembly, Berlin, Germany. The part concerning Earth rotation is mostly focused on the estimation of the geophysical excitation of polar motion using data from Gravity Recovery and Climate Experiment (GRACE) and its follow-on (GRACE-FO) missions, and on the improvement of the determination of Earth rotation parameters based on the Satellite Laser Ranging (SLR), Doppler Orbitography and Radiopositioning Integrated by Satellite (DORIS), and Global Navigation Satellite System (GNSS) satellite techniques. The part concerning geodynamics is focused on geodetic time series analysis for geodynamical purposes and monitoring of the vertical ground movements induced by mass transport within the Earth’s system, monitoring of the crustal movements using GNSS and newly applied Interferometric Synthetic Aperture Radar (InSAR), discussing the changes of the landslides and its monitoring using geodetic methods as well as investigations of seismic events and sea-level changes with geodetic methods. Finally, the recent research activities carried out by Polish scientists in the international projects is presented.

3

Forecasting municipal waste accumulation rate and personal consumption expenditures using vector autoregressive (VAR) model

Bień Jurand

Production Engineering Archives

|

2022

|

Vol. 28, Iss. 2

150--156

EN

Accurate forecasting of municipal solid waste (MSW) generation is important for the planning, operation and optimization of municipal waste management system. However, it’s not easy task due to dynamic changes in waste volume, its composition or unpredictable factors. Initially, mainly conventional and descriptive statistical models of waste generation forecasting with demographic and socioeconomic factors were used. Methods based on machine learning or artificial intelligence have been widely used in municipal waste projection for several years. This study investigates the trend of municipal waste accumulation rate and its relation to personal consumption expenditures based on the yearly data achieved from Local Data Bank (LDB) driven by Polish Statistical Office. The effect of personal consumption expenditures on the municipal waste accumulation rate was analysed by using the vector autoregressive model (VAR). The results showed that such method can be successfully used for this purpose with an approximate level of 2.3% Root Mean Square Error (RMSE).

4

The study of seismicity of Alborz (Northern Iran) and Zagros (Southern Iran) regions by using time series analysis

Rahimi Mohammad, Zamani Ahmad, Ghotbi Ali Reza

Acta Geophysica

|

2022

|

Vol. 70, no 1

27--37

EN

Since many years ago Iran faced multiple and sometimes destructive earthquakes because of its positions in seismic Alpidebelt, the regions of Alborz (in north of Iran) and Zagros (in south of Iran) are seismically important due to their dense population and the presence of economic and strategic organization in them, these zones with their high tectonic movements and their positions in seismic Alpide-belt have high potentials for large earthquakes, and 80% of Iran’s earthquakes happen in these zones. Time series analysis comprises methods for analyzing time series data in order to extract meaningful statistics and other characteristics of the data. Time series forecasting is the use of a model to predict future values based on previously observed values, in this study a proper model has been proposed for the earthquakes that occurred in two regions of Alborz and Zagros during 1900−2021, in all these regions time series are analyzed individually, and proper models are proposed for the prediction of empirical recurrence rate of the earthquake with magnitudes larger than 6.0 mb. Then, a compression was made between the proposed models and the earthquakes that occurred to determine their accuracy.

5

Wielomodelowe prognozowanie liczby pasażerów przewożonych krajowym transportem lotniczym w Stanach Zjednoczonych w 2021 r.

Kozicki Bartosz, Sowa Bogdan

Gospodarka Materiałowa i Logistyka

|

2021

|

nr 7

23--32

PL

W artykule przedstawiono dane dotyczące liczby pasażerów przewiezionych pasażerskim transportem lotniczym krajowym i międzynarodowym w Stanach Zjednoczonych w ujęciu miesięcznym w latach 2003–2020 i ich prognozowania na 2021 r. Badania rozpoczęto od analizy i oceny dwóch szeregów czasowych dotyczących liczby pasażerów przewiezionych transportem lotniczym pasażerskim w Stanach Zjednoczonych w ujęciu krajowym i międzynarodowym. Zbudowano model Kleina, za pomocą którego wykonano prognozowanie szeregu czasowego liczby pasażerów przewożonych transportem lotniczym krajowym w ujęciu miesięcznym na rok 2021. Zbudowany model jest połączeniem prognozowania ilościowego i jakościowego .

EN

The article presents data on the number of passengers transported by domestic and international passenger air transport in the United States on a monthly basis in the years 2003–2020 and their forecasting for 2021. The research began with the analysis and evaluation of two time series concerning the number of passengers transported by passenger air transport in the United States in terms of national and international approach. The Klein model was built, which was used to forecast the time series of the number of passengers transported by domestic air transport on a monthly basis for the year 2021. The constructed model is a combination of quantitative and qualitative forecasting.

6

A statistical study of COVID-19 pandemic in Egypt

Radwan Taha

Demonstratio Mathematica

|

2021

|

Vol. 54, nr 1

233--244

EN

The spread of the COVID-19 started in Wuhan on December 31, 2019, and a powerful outbreak of the disease occurred there. According to the latest data, more than 165 million cases of COVID-19 infection have been detected in the world (last update May 19, 2021). In this paper, we propose a statistical study of COVID-19 pandemic in Egypt. This study will help us to understand and study the evolution of this pandemic. Moreover, documenting of accurate data and taken policies in Egypt can help other countries to deal with this epidemic, and it will also be useful in the event that other similar viruses emerge in the future. We will apply a widely used model in order to predict the number of COVID-19 cases in the coming period, which is the autoregressive integrated moving average (ARIMA)model. This model depicts the present behaviour of variables through linear relationship with their past values. The expected results will enable us to provide appropriate advice to decision-makers in Egypt on how to deal with this epidemic.

7

Statistics of tropospheric amplitude scintillation over selected locations in tropical Nigeria

Ashidi Ayodeji Gabriel, Ojo Joseph Sunday, Kareem Afan Idowu, Ojo Olusola Samuel, Emmanuel Israel

Acta Geophysica

|

2021

|

Vol. 69, no. 3

947--957

EN

Tropospheric scintillation depends signifcantly on any location’s prevailing weather condition, and its variation must be statistically analyzed to ensure accurate fade margin determination. This study examines the distribution of Ku-band amplitude scintillation across selected locations in tropical Nigeria. Eight years of daily averaged data of surface temperature and relative humidity were employed for computing scintillation intensity (σ) and amplitude (χ) using international telecommunications union recommended model across eighteen (18) stations, that are subdivided into four (4) regions and spread over tropical Nigeria. The data, spanning January 2010 to December 2017, were obtained from the archive of the European center for medium-range weather forecasts (ECMWF) with a resolution of 0.125° by 0.125°. Three (3) years of in-situ data of concurrently measured satellite radio beacons and primary radio-climatic parameters at Akure (7° 17′ N, 5° 18′ E, 358 m), South-west Nigeria, were employed for comparison and validation. Statistical analyses involving time series, probability density, and cumulative distribution functions were performed on the scintillation dataset annually. Results indicate that the magnitude of tropospheric amplitude scintillation varies across diferent locations; nevertheless, it exhibits a similar distribution pattern characterized by the generalized extreme value (GEV) probability density function (pdf). The study has shown the need to incorporate the scintillation component into the fade mitigation architecture of telecommunication systems in tropical Nigeria while considering its regional variability. Also, experimental validation of the observations raised in this study should be encouraged at all the locations for better prediction accuracy.

8

Evaluation of the displacements of the permanent GPS stations of Beni Haroun Dam geodetic network (Algeria)

Khelifa Sofane, Dekkiche Hicham, Kahlouche Salem, Boudina Samia

Acta Geophysica

|

2021

|

Vol. 69, no. 6

2129--2143

EN

The purpose of this paper is the analysis of the daily coordinate time series of the five permanent GPS (Global Positioning System) stations of the geodetic monitoring network of the Beni-Haroun Dam (Algeria), in order to assess the spectral content of the dam displacements. The coordinate time series analysis was based on the singular spectrum analysis to assess their principal components (trend, seasonal components and noise in phase space), the spectral analysis to identify their noise spectrum (white or colored) and the wavelet thresholding method to determine their noise in frequency space. The results showed that the primary signal present in the analyzed time series is mainly composed of a trend and an annual component. The trend and the annual signal explain more than 95% of the total signal in the three coordinates (x, y, z) for all studied stations. The analyzed time series in the three coordinates (x, y, z) are characterized by a linear drift less than 1 mm/year, their annual amplitudes are in the range of 0.5–2 mm, and the amplitudes of their semiannual, four-monthly and quarterly signals are in the range of 0–0.5 mm. The noise spectrum in the analyzed time series is flicker noise, and the noise level is in the range of 0.2–0.7 mm, 0.3–0.5 mm and 0.5–1.2 mm in, respectively, x-, y- and z-coordinates. The low values of trend and noise level in the analyzed station coordinates indicate that the Beni-Haroun Dam is qualified as stable.

9

Effect of features extraction on improving LSTM network quality in ECG signal classification

Kłosowski Grzegorz

Przegląd Elektrotechniczny

|

2020

|

R. 96, nr 12

194--197

EN

This article focuses on the extraction of features extracted from ECG measurement signals to improve the quality of LSTM network operation. Two features were distinguished from each individual sequence of ECG signals: instantaneous frequency (IF) and spectral entropy (SE). Both of these features are extracted from ECG signals using short-time Fourier transform. The applied approach enables the conversion of original measurement sequences into spectral images, from which IF and SE coefficients are then generated. As a result of the research, it was found that feature extraction significantly improves ECG signal classification both in terms of forecasting accuracy and in terms of network learning speed.

PL

W niniejszym artykule skupiono się na ekstrakcji cech wyodrębnionych z sygnałów pomiarowych EKG w celu poprawy jakości działania sieci LSTM. Z każdej indywidualnej sekwencji sygnałów EKG wyróżniono dwie cechy: częstotliwość chwilową (IF) i entropię widmową (SE). Obie te cechy są wyodrębniane z sygnałów EKG przy użyciu krótkotrwałej transformaty Fouriera. Zastosowane podejście umożliwia konwersję oryginalnych sekwencji pomiarowych na obrazy widmowe, z których następnie generowane są współczynniki IF i SE. W wyniku badań stwierdzono, że ekstrakcja cech znacząco poprawia klasyfikację sygnału EKG zarówno pod względem dokładności prognozowania, jak i szybkości uczenia się.

10

Superiority of Simplicity: A Lightweight Model for Network Device Workload Prediction

Acker Alexander, Wittkopp Thorsten, Nedelkoski Sasho, Bogatinovski Jasmin, Kao Odej

Annals of Computer Science and Information Systems

|

2020

|

Vol. 21

7--10

EN

The rapid growth and distribution of IT systems increases their complexity and aggravates operation and maintenance. To sustain control over large sets of hosts and the connecting networks, monitoring solutions are employed and constantly enhanced. They collect diverse key performance indicators (KPIs) (e.g. CPU utilization, allocated memory, etc.) and provide detailed information about the system state. Storing such metrics over a period of time naturally raises the motivation of predicting future KPI progress based on past observations. This allows different ahead of time optimizations like anomaly detection or predictive maintenance. Predicting the future progress of KPIs can be defined as a time series forecasting problem. Although, a variety of time series forecasting methods exist, forecasting the progress of IT system KPIs is very hard. First, KPI types like CPU utilization or allocated memory are very different and hard to be modelled by the same model. Second, system components are interconnected and constantly changing due to soft- or firmware updates and hardware modernization. Thus a frequent model retraining or fine-tuning must be expected. Therefore, we propose a lightweight solution for KPI series prediction based on historic observations. It consists of a weighted heterogeneous ensemble method composed of two models - a neural network and a mean predictor. As ensemble method a weighted summation is used, whereby a heuristic is employed to set the weights. The lightweight nature allows to train models individually on each KPI series and makes model retraining feasible when system changes occur. The modelling approach is evaluated on the available FedCSIS 2020 challenge dataset and achieves an overall R^2 score of 0.10 on the preliminary 10\% test data and 0.15 on the complete test data. We publish our code on the following github repository: https://github.com/citlab/fed\_challenge.

11

Separation of split shear waves based on a hodogram analysis of HTI media

Yang Y., Lu J., Wang Y.

Acta Geophysica

|

2018

|

Vol. 66, no. 4

643--658

EN

Although the shear-wave birefringence phenomenon affects the imaging of converted shear waves, it also provides a considerable amount of information on subsurface fracture development. Therefore, it is significant to separate split shear waves before seismic interpretation and reservoir prediction. In this paper, we propose a new method of split shear waves separation based on the polarization directions derived from hodogram analysis. Through the hodogram analysis, we find that the split shear-wave particle motions are within the range of a specific and fixed rectangle, which have relations with the fracture azimuth in strata. In addition, we found that a couple of split shear waves can only be fitted to the unique trajectory rectangle through the theoretical derivation. Based on this, we establish the trajectory rectangle through the wave vector calculation and calculate the fracture azimuth according to the fact that the one edge of the trajectory rectangle is along or perpendicular to the fracture azimuth. Synthetic data analysis shows that the calculation accuracy of fracture azimuth under the constraint of trajectory rectangle is less affected by the time delay between split shear waves than using the method of eigenvector–eigenvalue decomposition (EED). Therefore, we can obtain better results for separation of split shear waves using our method than using EED. Eventually, we propose an approach of layer stripping to deal with the problem that shear wave split several times due to the situation that different strata have different fracture azimuths. Synthetic data test indicates that our method can achieve higher calculation efficiency and faster convergence speed than the conventional eigenvector–eigenvalue decomposition method, even though the data are of a low signal-to-noise ratio. Moreover, field data applications show the effectiveness and potential of our method.

12

Zastosowanie metod eksploracji danych produkcyjnych w diagnostyce przyczyn wad wyrobów i zakłóceń procesu produkcyjnego

Perzyk M.

Utrzymanie Ruchu

|

2017

|

Nr 4

45--47

PL

Metody eksploracji danych mogą przynieść znaczące korzyści w procesach produkcyjnych, przyczyniając się do ułatwienia wykrywania przyczyn problemów w postaci wad wyrobów i innych zakłóceń procesów wytwarzania. Warto zwrócić uwagę, że metody te z założenia wykorzystują istniejące, zarejestrowane w przedsiębiorstwie dane, bez konieczności przeprowadzania kosztownych eksperymentów w warunkach laboratoryjnych lub przemysłowych. Warunkiem pomyślnego ich stosowania jest jednak uświadomienie sobie przez personel inżynieryjny ogromnego potencjału systemów eksploracji danych w przedsiębiorstwach produkcyjnych, których wprowadzenie stanie się niedługo koniecznością.

13

Real time search algorithm for observation outliers during monitoring engineering constructions

Latos D., Kolanowski B., Pachelski W., Sołoducha R.

Reports on Geodesy and Geoinformatics

|

2017

|

Vol. 104

103--114

EN

Real time monitoring of engineering structures in case of an emergency of disaster requires collection of a large amount of data to be processed by specific analytical techniques. A quick and accurate assessment of the state of the object is crucial for a probable rescue action. One of the more significant evaluation methods of large sets of data, either collected during a specified interval of time or permanently, is the time series analysis. In this paper presented is a search algorithm for those time series elements which deviate from their values expected during monitoring. Quick and proper detection of observations indicating anomalous behavior of the structure allows to take a variety of preventive actions. In the algorithm, the mathematical formulae used provide maximal sensitivity to detect even minimal changes in the object’s behavior. The sensitivity analyses were conducted for the algorithm of moving average as well as for the Douglas-Peucker algorithm used in generalization of linear objects in GIS. In addition to determining the size of deviations from the average it was used the so-called Hausdorff distance. The carried out simulation and verification of laboratory survey data showed that the approach provides sufficient sensitivity for automatic real time analysis of large amount of data obtained from different and various sensors (total stations, leveling, camera, radar).

14

Changes in seasonal streamflow extremes experienced in rivers of Northwestern South America (Colombia)

Pierini J.O., Restrepo J.C., Aguirre J., Bustamante A.M., Velásquez G.J.

Acta Geophysica

|

2017

|

Vol. 65, no. 2

377--394

EN

A measure of the variability in seasonal extreme streamflow was estimated for the Colombian Caribbean coast, using monthly time series of freshwater discharge from ten watersheds. The aim was to detect modifications in the streamflow monthly distribution, seasonal trends, variance and extreme monthly values. A 20-year length time moving window, with 1-year successive shiftments, was applied to the monthly series to analyze the seasonal variability of streamflow. The seasonal-windowed data were statistically fitted through the Gamma distribution function. Scale and shape parameters were computed using the Maximum Likelihood Estimation (MLE) and the bootstrap method for 1000 resample. A trend analysis was performed for each windowed-serie, allowing to detect the window of maximum absolute values for trends. Significant temporal shifts in seasonal streamflow distribution and quantiles (QT), were obtained for different frequencies. Wet and dry extremes periods increased significantly in the last decades. Such increase did not occur simultaneously through the region. Some locations exhibited continuous increases only at minimum QT.

15

Techniki budowy nieliniowych modeli diagnostycznych z wykorzystaniem szeregów czasowych drgań mechanicznych

Puchalski A.

Przegląd Mechaniczny

|

2016

|

nr 10

33--36

PL

W artykule omówiono zastosowanie metod analizy nieliniowych szeregów czasowych drgań do oceny ich wartości informacyjnej do celów diagnostycznych układów mechanicznych na przykładzie zespołów napędowych w pojazdach. Trzy techniki detekcji oraz ekstrakcji sygnałów obejmowały analizę w zrekonstruowanej, zanurzonej przestrzeni fazowej szeregu czasowego, prowadzącą do segmentacji sygnału w zakresie dynamiki chaotycznej i losowej, analizę multifraktalną oraz analizę rozkładów stabilnych prawdopodobieństwa.

EN

The paper discusses the use of methods of nonlinear time series analysis of vibrations to assess their informational value for the diagnostic purposes of the mechanical systems on the example of vehicle powertrain. Three detection and extraction techniques of signals included an analysis of the reconstructed embedded phase space of time series which leads to the segmentation of the signal in terms of the chaotic and random dynamics, multifractal analysis and the analysis of alpha-stable probability distributions.

16

Application of Time-Series Analysis for Predicting Defects in Continuous Steel Casting Process

Rodziewicz A., Perzyk M.

Archives of Foundry Engineering

|

2016

|

Vol. 16, iss. 4

125--130

EN

The purpose of this paper was testing suitability of the time-series analysis for quality control of the continuous steel casting process in production conditions. The analysis was carried out on industrial data collected in one of Polish steel plants. The production data concerned defective fractions of billets obtained in the process. The procedure of the industrial data preparation is presented. The computations for the time-series analysis were carried out in two ways, both using the authors’ own software. The first one, applied to the real numbers type of the data has a wide range of capabilities, including not only prediction of the future values but also detection of important periodicity in data. In the second approach the data were assumed in a binary (categorical) form, i.e. the every heat(melt) was labeled as ‘Good’ or ‘Defective’. The naïve Bayesian classifier was used for predicting the successive values. The most interesting results of the analysis include good prediction accuracies obtained by both methodologies, the crucial influence of the last preceding point on the predicted result for the real data time-series analysis as well as obtaining an information about the type of misclassification for binary data. The possibility of prediction of the future values can be used by engineering or operational staff with an expert knowledge to decrease fraction of defective products by taking appropriate action when the forthcoming period is identified as critical.

17

Wykorzystanie modelu ARIMA do analizy szeregu czasowego

Malska W., Wachta H.

Zeszyty Naukowe Politechniki Rzeszowskiej. Elektrotechnika

|

2015

|

z. 34 [292], nr 3

23--30

PL

W artykule zaprezentowano zastosowanie metody ARIMA służącej do analizy szeregu czasowego z trendem i sezonowością. Szereg czasowy jest jednym z rodzajów szeregów statystycznych, który można zdefiniować jako ciąg obserwacji pewnego zjawiska w kolejnych jednostkach czasu (latach, kwartałach, miesiącach, itp.). Analiza szeregów czasowych opiera się na głównym założeniu, że kolejne wartości rozważanej cechy (zmiennej) reprezentują kolejne pomiary wykonane w takiej samej jednostce czasu (w równych odstępach czasu). Zmienną niezależną jest czas (jednostka czasu). Obserwując różne zjawiska (w tym także związane z gospodarką elektroenergetyczną) często chcemy wiedzieć czy i jak zmieniają się w czasie, czyli jaka jest ich dynamika. Analiza szeregów czasowych stosowana jest głównie do podejmowania decyzji związanych z przyszłością. Rozważane zjawisko może podlegać pewnym prawidłowościom, których wykrycie i opis jest głównym celem analizy szeregów czasowych. W wielu przypadkach modele szeregów czasowych wykorzystywane są w celu wnioskowania o przyszłości badanego zjawiska (do prognozowania). Prognozowanie w ujęciu statystycznym to wnioskowanie o przyszłych wartościach szeregu czasowego, które oparte jest na danych czasowych lub analizie wartości, jakie przyjmują rozważane cechy statystyczne (zmienne). Przy analizie w dziedzinie czasu w szeregu czasowym można wyodrębnić pewne składowe (stały przeciętny poziom zjawiska, trend, cykle długookresowe, wahania sezonowe, wahania krótkookresowe, interwencje, składnik losowy (zakłócenie losowe)), przy czym nie wszystkie one muszą występować w konkretnym analizowanym szeregu. Metoda prognozowania zależy od składowych szeregu czasowego. Wyniki obliczeń z wykorzystaniem modelu ARIMA zaprezentowano korzystając z pakietu STATISTICA v. 10.0.

EN

The paper presents the application of the method used for the analysis of ARIMA time series with trend and seasonality. Time series is one of the types of statistical series, which can be defined as a series of observations of a phenomenon in the following units of time (years, quarters, months, etc.). Time series analysis based on the main idea that a further consideration of the characteristics (variable) represent the more measurements made in the same unit of time (at regular intervals). The independent variable is the time (unit of time). Observing different phenomena (including related to the economy electricity) often want to know whether and how they are changing over time, that is what is their dynamics. Time series analysis is mainly used to make decisions about the future. Considered phenomenon may be subject to certain regularities, which detect and description is the main objective of the analysis of time series. In many cases, time series models are used to apply for the future of the studied phenomenon (to predict). Forecasting is statistically inference about future values of the time series, which is based on the analysis of data or time values which take under consideration the statistical characteristics (variables). At the time domain analysis in time series can extract some components (constant average level of the phenomenon, a trend long-term cycles, seasonal fluctuations, fluctuations in short-term, interventions, random component (random disturbance)), and not all of them must be analyzed in a specific number of. Forecasting method depends on the components of the time series. The results of calculations using the ARIMA model is presented using STATISTICA v. 10.0.

18

Krótkoterminowe prognozowanie wielkości popytu z wykorzystaniem klasycznych metod predykcji na przykładzie międzynarodowego przedsiębiorstwa produkcyjno-dystrybucyjnego

Jurczyk K., Gdowska K.

Logistyka

|

2015

|

nr 4

9076--9084, CD3

PL

W pracy zaprezentowano wyniki badań polegających na doborze odpowiedniej metody do prognozowania wielkości popytu w międzynarodowym przedsiębiorstwie produkcyjno-dystrybucyjnym. Wykorzystane metody – naiwną, średniej ruchomej, wygładzania wykładniczego oraz wskaźników sezonowości – porównano ze sobą oraz zwrócono wagę co do zasadności ich stosowania. Obiektem badań było międzynarodowe przedsiębiorstwo produkcyjno-dystrybucyjne, którego nazwa została zakodowana jako Przedsiębiorstwo X. W badaniach przeanalizowano wielkość sprzedaży ośmiu wybranych produktów na rynku skandynawskim. Historyczne miesięczne dane sprzedaży pochodziły z lat 2005–2011.

EN

In the paper a short-term demand forecasting models for international production and distribution enterprise were compared. The object of analysis was an international production and distribution enterprise which brand name was coded as the “X Enterprise”. In the research eight chosen products’ sales volumes were analysed. Historical input data came from 2005–2011.

19

Forecasting European thermal coal spot prices

Krzemień A., Riesgo Fernandez P., Suárez Sánchez A., Sánchez Lasheras F.

Journal of Sustainable Mining

|

2015

|

Vol. 14, iss. 4

203--210

EN

This paper presents a one-year forecast of European thermal coal spot prices by means of time series analysis, using data from IHS McCloskey NW Europe Steam Coal marker (MCIS). The main purpose was to achieve a good fit for the data using a quick and feasible method and to establish the transformations that better suit this marker, together with an affordable way for its validation. Time series models were selected because the data showed an autocorrelation systematic pattern and also because the number of variables that influence European coal prices is very large, so forecasting coal prices as a dependent variable makes necessary to previously forecast the explanatory variables. A second-order Autoregressive process AR(2) was selected based on the autocorrelation and the partial autocorrelation function. In order to determine if the results obtained are a good fit for the data, the possible drivers that move the European thermal coal spot prices were taken into account, establishing a hypothesis in which they were divided into four categories: (1) energy side drivers, that directly relates coal prices with other energy commodities like oil and natural gas; (2) demand side drivers, that relates coal prices both with the Western World economy and with emerging economies like China, in connection with the demand for electricity in these economies; (3) commodity currency drivers, that have an influence for holders of different commodity currencies in countries that export or import coal; and (4) supply side drivers, involving the production costs, transportation, etc. Finally, in order to analyse the time series model performance a Generalized Regression Neural Network (GRNN) was used and its performance compared against the whole AR(2) process. Empirical results obtained confirmed that there is no statistically significant difference between both methods. The GRNN analysis also allowed pointing out the main drivers that move the European Thermal Coal Spot prices: crude oil, USD/CNY change and supply side drivers.

20

Application of Special Cause Control Charts to Green Sand Process

Perzyk M., Rodziewicz A.

Archives of Foundry Engineering

|

2015

|

Vol. 15, iss. 4

55--60

EN

Statistical Process Control (SPC) based on the well known Shewhart control charts, is widely used in contemporary manufacturing industry, including many foundries. However, the classic SPC methods require that the measured quantities, e.g. process or product parameters, are not auto-correlated, i.e. their current values do not depend on the preceding ones. For the processes which do not obey this assumption the Special Cause Control (SCC) charts were proposed, utilizing the residual data obtained from the time-series analysis. In the present paper the results of application of SCC charts to a green sand processing system are presented. The tests, made on real industrial data collected in a big iron foundry, were aimed at the comparison of occurrences of out-of-control signals detected in the original data with those appeared in the residual data. It was found that application of the SCC charts reduces numbers of the signals in almost all cases It is concluded that it can be helpful in avoiding false signals, i.e. resulting from predictable factors.