Wyniki wyszukiwania - BazTech

1

Cuban consumer price index forecasting through transformer with attention

Rosado Reynaldo, Toledano-López Orlando G., González Hector R., Abreu Aldis J., Hernandez Yanio

Journal of Automation Mobile Robotics and Intelligent Systems

|

2023

|

Vol. 17, No. 2

12--17

EN

Recently, time series forecasting modelling in the Con‐ sumer Price Index (CPI) has attracted the attention of the scientific community. Several research projects have tackled the problem of CPI prediction for their countries using statistical learning, machine learning and deep neural networks. The most popular approach to CPI in several countries is the Autoregressive Integrated Mov‐ ing Average (ARIMA) due to the nature of the data. This paper addresses the Cuban CPI forecasting problem using Transformer with attention model over univariate dataset. The fine tuning of the lag parameter shows that Cuban CPI has better performance with small lag and that the best result was in 𝑝 = 1. Finally, the comparative results between ARIMA and our proposal show that the Transformer with attention has a very high performance despite having a small data set.

2

Impact of starting outlier removal on accuracy of time series forecast ING

Romanuke Vadim

Maritime Technical Journal

|

2022

|

nr 1 (224)

1--15

EN

The presence of an outlier at the starting point of a univariate time series negatively influences the forecasting accuracy. The starting outlier is effectively removed only by making it equal to the second time point value. The forecasting accuracy is significantly improved after the removal. The favorable impact of the starting outlier removal on the time series forecasting accuracy is strong. It is the least favorable for time series with exponential rising. In the worst case of a time series, on average only 7 % to 11 % forecasts after the starting outlier removal are worse than they would be without the removal.

PL

Wartość odstająca w punkcie początkowym jednowymiarowego szeregu czasowego negatywnie wpływa na dokładność prognozowania. W ramach przeprowadzonych badań dokonano analizy wpływu usunięcia wartości odstającej poprzez zrównanie jej z wartością drugiego punktu cza-sowego. Uzyskane wyniki wskazują, że przyjęta metoda znacznie poprawia dokładność progno-zowania dla większości szeregów czasowych. Jednak w przypadku szeregów czasowych z wykładniczym wzrostem, metoda okazała się mniej skuteczna. Minimalny wzrost dokładności prognozowania wynosił w tym przypadku od 7 % do 11 %.

3

Arima Model optimal selection for time series forecasting

Romanuke Vadim

Maritime Technical Journal

|

2022

|

nr 1 (224)

28--40

EN

A fast-and-flexible method of ARIMA model optimal selection is suggested for univariate time series forecasting. The method allows obtaining as-highly-accurate-as-possible forecasts auto-matically. It is based on effectively finding lags by the autocorrelation function of a detrended time series, where the best-fitting polynomial trend is subtracted from the time series. The fore-casting quality criteria are the root-mean-square error (RMSE) and the maximum absolute error (MaxAE) allowing to register information about the average inaccuracy and worst outlier. Thus, the ARIMA model optimal selection is performed by simultaneously minimizing RMSE and Max-AE, whereupon the minimum defines the best model. Otherwise, if the minimum does not exist, a combination of minimal-RMSE and minimal-MaxAE ARIMA models is used.

PL

W pracy zaproponowano szybką i elastyczną metodę optymalnego doboru modelu ARIMA na potrzeby prognozowania szeregów czasowych z jedną zmienną. Metoda pozwala na uzyskanie możliwie najdokładniejszych prognoz, opierając się na skutecznym znajdowaniu opóźnień. Po-szukiwanie opóźnień realizowane jest za pomocą funkcji autokorelacji szeregu czasowego bez trendu, w którym najlepiej dopasowany trend wielomianowy jest odejmowany od szeregu cza-sowego. Za kryteria jakości prognozowania przyjęto średni błąd kwadratowy (RMSE) i maksy-malny błąd bezwzględny (MaxAE), które pozwoliły na rejestrację informacji o średniej i maksymalnej niedokładności. Optymalny dobór modelu ARIMA odbywa się poprzez jednoczesną minimalizację RMSE i MaxAE, dla której wartość minimalna określa najlepszy model. W przeciw-nym razie, jeśli minimum nie istnieje, używana jest kombinacja modeli ARIMA z minimalnym RMSE i minimalnym MaxAE.

4

Data analysis-based time series forecast for managing household electricity consumption

Bezzar Nour El-Houda, Laimeche Lakhdar, Meraoumia Abdallah, Houam Lotfi

Demonstratio Mathematica

|

2022

|

Vol. 55, nr 1

900--921

EN

Recently, electricity consumption forecasting has attracted much research due to its importance in our daily life as well as in economic activities. This process is seen as one of the ways to manage future electricity needs, including anticipating the supply-demand balance, especially at peak times, and helping the customer make real-time decisions about their consumption. Therefore, based on statistical techniques (ST) and/or artificial intelligence (AI), many forecasting models have been developed in the literature, but unfortunately, in addition to poor choice of the appropriate model, time series datasets were used directly without being seriously analyzed. In this article, we have proposed an efficient electricity consumption prediction model that takes into account the shortcomings mentioned earlier. Therefore, the database was analyzed to address all anomalies such as non-numeric values, aberrant, and missing values. In addition, by analyzing the correlation between the data, the possible periods for forecasting electricity consumption were determined. The experimental results carried out on the Individual Household Electricity Power Consumption dataset showed a clear superiority of the proposed model over most of the ST and/or AI-based models proposed in the literature.

5

Applying Python’s Time Series Forecasting Method in Microsoft Excel – Integration as a Business Process Supporting Tool for Small Enterprises

Litwin Jolanta, Olech Marcin, Szymusik Anna

Technical Sciences / University of Warmia and Mazury in Olsztyn

|

2021

|

nr 24(1)

115--133

EN

The paper describes the current state of research, where integration of Microsoft Excel and Python interpreter, gives the business user the right tool to solve chosen business process analysis problems like: forecasting, classification or clustering. The integration is done by using Visual Basic for Application (VBA), as well as XLWings Python’s library. Both mechanisms serve as an interfaces between MS Excel and Python to allow the data exchange between each other. Creating the suitable Graphical User Interface (GUI) in Microsoft Excel, gives the business user opportunity to select specific data analysis method available in Python’s environment and set its parameters, without Python’s programming. Running the method by Python’s interpreter can bring the results, which are hard or even impossible to obtain by using Microsoft Excel only. However, the data analysis methods stored in the Python’s script, which are available to the business user, as well as VBA source code, must be designed and implemented by the data scientist. Sample, basic integration between Microsoft Excel and Python’s interpreter is presented in the paper. To present value-added of the proposed software solution, simple case study according to time series forecasting problem is described, where forecasting errors of different methods available in the Microsoft Excel and Python are presented and discussed. The paper ends with conclusions according to the results of the current researches and suggested directions of further research.

6

Time Series Forecasting with Data Transform and Its Application in Sport

Do Hao, Chau Duc, Cai Thuc, Lam Han, Nguyen Dat, Lam Tuong, Tran Son

Annals of Computer Science and Information Systems

|

2021

|

Vol. 27

29--32

EN

Forecasting time series data is an exciting challenge. Although being complex, this is a high potential for industrial use. One of the most significant gaps in the forecasting process is the quality of data representation, especially with the time-series data. This paper proposes an effective method using an integral transform that can show hidden information of the time series data. The integral transform exploits data as a composition of many basic functions and then use this set to present the data. Mathematically, this transform converts the data into another space with another feature, showing many properties hidden in the original form. The experimental result demonstrates our suggestion can learn the transformation rules and then can be applied for many applications.

7

Continuous blood glucose level prediction of Type 1 Diabetes based on Artificial Neural Network

Ben Ali J., Hamdi T., Fnaiech N., Di Costanzo V., Fnaiech F., Ginoux J. M.

Biocybernetics and Biomedical Engineering

|

2018

|

Vol. 38, no. 4

828--840

EN

Recent technological advancements in diabetes technologies, such as Continuous Glucose Monitoring (CGM) systems, provide reliable sources to blood glucose data. Following its development, a new challenging area in the field of artificial intelligence has been opened and an accurate prediction method of blood glucose levels has been targeted by scientific researchers. This article proposes a new method based on Artificial Neural Networks (ANN) for blood glucose level prediction of Type 1 Diabetes (T1D) using only CGMdata as inputs. To show the efficiency of our method and to validate our ANN, real CGM data of 13 patients were investigated. The accuracy of the strategy is discussed based on some statistical criteria such as the Root Mean Square Error (RMSE) and the Mean Absolute Percentage Error (MAPE). The obtained averages of RMSE are 6.43 mg/dL, 7.45 mg/dL, 8.13 mg/dL and 9.03 mg/dL for Prediction Horizon (PH) respectively 15 min, 30 min, 45 min and 60 min and the average of MAPE was 3.87% for PH = 15 min, knowing that the smaller is the RMSE and MAPE, the more accurate is the prediction. Experimental results show that the proposed ANN is accurate, adaptive, and very encouraging for a clinical implementation. Furthermore, while other studies have only focused on the prediction accuracy of blood glucose, this work aims to improve the quality of life of T1D patients by using only CGM data as inputs and by limiting human intervention.

8

Accurate prediction of continuous blood glucose based on support vector regression and differential evolution algorithm

Hamdi T., Ben Ali J., Di Costanzo V., Fnaiech F., Moreau E., Ginoux J. M.

Biocybernetics and Biomedical Engineering

|

2018

|

Vol. 38, no. 2

362--372

EN

Type 1 diabetes (T1D) is a chronic disease requiring patients to know their blood glucose values in order to ensure blood glucose levels as close to normal as possible. Hence, the ability to predict blood glucose levels is of a great interest for clinical researchers. In this sense, the literature is rich with several solutions that can predict blood glucose levels. Unfortunately, these methods require the patient to specific their daily activities: meal intake, insulin injection and emotional factors, which can be error prone. To reduce this burden on the patent, this work proposes to use only continuous glucose monitoring (CGM) data to predict blood glucose levels independently of other factors. To support this, support vector regression (SVR) and differential evolution (DE) algorithms were investigated. The proposed method is validated using real CGM data of 12 patients. The obtained average of root mean square error (RMSE) was 9.44, 10.78, 11.82 and 12.95 mg/dL for prediction horizon (PH) respectively equal to 15, 30, 45 and 60 min. The results of the present study and comparison with some previous works show that the proposed method holds promise. The SVR based on DE algorithm achieved high prediction accuracy while being robustness, automatic, and requiring no human intervention.

9

Predictability of monthly temperature and precipitation using automatic time series forecasting methods

Papacharalampous G., Tyralis H., Koutsoyiannis D.

Acta Geophysica

|

2018

|

Vol. 66, no. 4

807--831

EN

We investigate the predictability of monthly temperature and precipitation by applying automatic univariate time series forecasting methods to a sample of 985 40-year-long monthly temperature and 1552 40-year-long monthly precipitation time series. The methods include a naïve one based on the monthly values of the last year, as well as the random walk (with drift), AutoRegressive Fractionally Integrated Moving Average (ARFIMA), exponential smoothing state-space model with Box–Cox transformation, ARMA errors, Trend and Seasonal components (BATS), simple exponential smoothing, Theta and Prophet methods. Prophet is a recently introduced model inspired by the nature of time series forecasted at Facebook and has not been applied to hydrometeorological time series before, while the use of random walk, BATS, simple exponential smoothing and Theta is rare in hydrology. The methods are tested in performing multi-step ahead forecasts for the last 48 months of the data. We further investigate how different choices of handling the seasonality and non-normality affect the performance of the models. The results indicate that: (a) all the examined methods apart from the naïve and random walk ones are accurate enough to be used in long-term applications; (b) monthly temperature and precipitation can be forecasted to a level of accuracy which can barely be improved using other methods; (c) the externally applied classical seasonal decomposition results mostly in better forecasts compared to the automatic seasonal decomposition used by the BATS and Prophet methods; and (d) Prophet is competitive, especially when it is combined with externally applied classical seasonal decomposition.

10

An artificial neural network based on time series prediction system with distributed computing

Lisowski M.

Zeszyty Naukowe Akademii Morskiej w Gdyni

|

2013

|

nr 79

32--40

EN

In section 1, some aspects of time series forecasting are introduced. Two subsequent sections briefly discuss the application of multi-layer feed-forward neural networks for making predictions based on past data, and the automation of this process. The paper is concluded with a summary of an experiment consisting of running an implementation of an automated ANN based prediction system on examples of real life time series data, and basic conclusions drawn from some of the results of this experiment.

11

Wspomaganie planowania wielkości zapotrzebowania na klej poliuretanowy w kopalni węgla kamiennego

Jakowska-Suwalska K., Sojda A., Wolny M.

Zeszyty Naukowe. Organizacja i Zarządzanie / Politechnika Śląska

|

2012

|

z. 61

127--138

PL

Praca przedstawia propozycję metody wspomagania planowania zapotrzebowania na klej poliuretanowy, która bazuje na metodach prognozowania szeregów czasowych oraz na podstawie modelu ekonometrycznego. Jako finalny model prognostyczny wspomagający planowanie wielkości zapotrzebowania zaproponowano kombinowany model agregujący prognozy postawione za pomocą wybranych modeli. Agregacja polega na zastosowaniu sumy ważonej, przy tym wagi ustalono na podstawie kryterium minimalnego błędu prognoz wygasłych.

EN

In this paper proposal of method for polyurethane adhesive demand planning support is presented. The method is based on models of time series forecasting and econometric model. The proposal is to combine the forecasts through application of weighted sum. The weight factors are determined by the minimal mean error of extinct forecasts criterion.

12

Intelligent financial time series forecasting: A complex neuro-fuzzy approach with multi-swarm intelligence

Li C., Chiang T. W.

International Journal of Applied Mathematics and Computer Science

|

2012

|

Vol. 22, no. 4

787-800

EN

Financial investors often face an urgent need to predict the future. Accurate forecasting may allow investors to be aware of changes in financial markets in the future, so that they can reduce the risk of investment. In this paper, we present an intelligent computing paradigm, called the Complex Neuro-Fuzzy System (CNFS), applied to the problem of financial time series forecasting. The CNFS is an adaptive system, which is designed using Complex Fuzzy Sets (CFSs) whose membership functions are complex-valued and characterized within the unit disc of the complex plane. The application of CFSs to the CNFS can augment the adaptive capability of nonlinear functional mapping, which is valuable for nonlinear forecasting. Moreover, to optimize the CNFS for accurate forecasting, we devised a new hybrid learning method, called the HMSPSO-RLSE, which integrates in a hybrid way the so-called Hierarchical Multi-Swarm PSO (HMSPSO) and the well known Recursive Least Squares Estimator (RLSE). Three examples of financial time series are used to test the proposed approach, whose experimental results outperform those of other methods.

13

Improving prediction models applied in systems monitoring natural hazards and machinery

Sikora M., Sikora B.

International Journal of Applied Mathematics and Computer Science

|

2012

|

Vol. 22, no. 2

477-491

EN

A method of combining three analytic techniques including regression rule induction, the k-nearest neighbors method and time series forecasting by means of the ARIMA methodology is presented. A decrease in the forecasting error while solving problems that concern natural hazards and machinery monitoring in coal mines was the main objective of the combined application of these techniques. The M5 algorithm was applied as a basic method of developing prediction models. In spite of an intensive development of regression rule induction algorithms and fuzzy-neural systems, the M5 algorithm is still characterized by the generalization ability and unbeatable time of data model creation competitive with other systems. In the paper, two solutions designed to decrease the mean square error of the obtained rules are presented. One consists in introducing into a set of conditional variables the so-called meta-variable (an analogy to constructive induction) whose values are determined by an autoregressive or the ARIMA model. The other shows that limitation of a data set on which the M5 algorithm operates by the k-nearest neighbor method can also lead to error decreasing. Moreover, three application examples of the presented solutions for data collected by systems of natural hazards and machinery monitoring in coal mines are described. In Appendix, results of several benchmark data sets analyses are given as a supplement of the presented results.

14

Heuristic modeling using recurrent neural networks: simulated and real-data experiments

Przystałka P.

Computer Assisted Mechanics and Engineering Sciences

|

2007

|

Vol. 14, No. 4

715-727

EN

The focus of this paper is on the problems of system identification, process modeling and time series forecasting which can be met during the use of locally recurrent neural networks in heuristic modeling technique. However, the main interest of this paper is to survey the properties of the dynamic neural processor which is developed by the author. Moreover, a comparative study of selected recurrent neural architectures in modeling tasks is given. The results of experiments showed that some processes tend to be chaotic and in some cases it is reasonable to use soft computing models for fault diagnosis and control.