scholarly journals Traffic Speed Prediction Based on Heterogeneous Graph Attention Residual Time Series Convolutional Networks

AI ◽  
2021 ◽  
Vol 2 (4) ◽  
pp. 650-661
Author(s):  
Yan Du ◽  
Xizhong Qin ◽  
Zhenhong Jia ◽  
Kun Yu ◽  
Mengmeng Lin
Keyword(s):  
Time Series ◽  
Traffic Accidents ◽  
Traffic Prediction ◽  
Traffic Data ◽  
Unified Framework ◽  
Real World Data ◽  
Convolutional Network ◽  
Residual Time ◽  
Social Events

Accurate and timely traffic forecasting is an important task for the realization of urban smart traffic. The random occurrence of social events such as traffic accidents will make traffic prediction particularly difficult. At the same time, most of the existing prediction methods rely on prior knowledge to obtain traffic maps and the obtained map structure cannot be guaranteed to be accurate for the current learning task. In addition, traffic data is highly non-linear and long-term dependent, so it is more difficult to achieve accurate prediction. In response to the above problems, this paper proposes a new integrated unified architecture for traffic prediction based on heterogeneous graph attention network combined with residual-time-series convolutional network, which is called HGA-ResTCN. First, the heterogeneous graph attention is used to capture the changes in the relationship between the traffic graph nodes caused by social events, so as to learn the link weights between the target node and its neighbor nodes; at the same time, by introducing the timing of residual links convolutional network to capture the long-term dependence of complex traffic data. These two models are integrated into a unified framework to learn in an end-to-end manner. Through testing on real-world data sets, the results show that the accuracy of the model in this paper is better than other proposed baselines.

scholarly journals OPTIMISING THE SMOOTHNESS AND ACCURACY OF MOVING AVERAGE FOR STOCK PRICE DATA

2018 ◽  
Vol 24 (3) ◽  
pp. 984-1003 ◽  
Author(s):  
Aistis RAUDYS ◽  
Židrina PABARŠKAITĖ
Keyword(s):  
Time Series ◽  
Real World ◽  
Stock Price ◽  
Weather Forecasting ◽  
Moving Average ◽  
New Method ◽  
Traffic Prediction ◽  
Real World Data ◽  
Price Series ◽  
Unseen Data

Smoothing time series allows removing noise. Moving averages are used in finance to smooth stock price series and forecast trend direction. We propose optimised custom moving average that is the most suitable for stock time series smoothing. Suitability criteria are defined by smoothness and accuracy. Previous research focused only on one of the two criteria in isolation. We define this as multi-criteria Pareto optimisation problem and compare the proposed method to the five most popular moving average methods on synthetic and real world stock data. The comparison was performed using unseen data. The new method outperforms other methods in 99.5% of cases on synthetic and in 91% on real world data. The method allows better time series smoothing with the same level of accuracy as traditional methods, or better accuracy with the same smoothness. Weights optimised on one stock are very similar to weights optimised for any other stock and can be used interchangeably. Traders can use the new method to detect trends earlier and increase the profitability of their strategies. The concept is also applicable to sensors, weather forecasting, and traffic prediction where both the smoothness and accuracy of the filtered signal are important.

The Long-Term Wind and Wave Spectra of the Mediterranean Sea

2005 ◽  
Author(s):  
Christos N. Stefanakos
Keyword(s):  
Time Series ◽  
Standard Deviation ◽  
Stochastic Processes ◽  
Mediterranean Sea ◽  
Mean Value ◽  
Residual Time ◽  
Wave Data ◽  
The Mediterranean ◽  
Periodic Components

It is a well-known fact that long-term time series of wind and wave data are modelled as nonstationary stochastic processes with yearly periodic mean value and standard deviation (periodically correlated or cyclostationary stochastic processes). Using this model, the initial nonstationary series are decomposed to a seasonal (periodic) mean value m(t) and a residual time series W(t) multiplied by a seasonal (periodic) standard deviation s(t), of the form Y(t) = m(t) + s(t)W(t). The periodic components m(t) and s(t) are estimated using mean monthly values, and the residual time series W(t) is examined for stationarity. For this purpose, spectral densities of W(t) are obtained from different seasonal segments, calculated and compared with each other. It is shown that W(t) can indeed be considered stationary, and thus Y(t) can be considered periodically correlated. This analysis has been applied to model wind and wave data from several locations in the Mediterranean Sea. It turns out that the spectrum of W(t) is very weakly dependent on the site, a fact that might be useful for the geographic parameterization of wind and wave climate.

scholarly journals A Robust Data-Driven Method for Multiseasonality and Heteroscedasticity in Time Series Preprocessing

2021 ◽  
Vol 2021 ◽  
pp. 1-11
Author(s):  
Bin Sun ◽  
Liyao Ma ◽  
Tao Shen ◽  
Renkang Geng ◽  
Yuan Zhou ◽  
...  
Keyword(s):  
Time Series ◽  
Operating Characteristic ◽  
Measured Data ◽  
Data Sources ◽  
Data Driven ◽  
Data Transport ◽  
Real World Data ◽  
Promising Solution ◽  
Labelling Method

Internet of Things (IoT) is emerging, and 5G enables much more data transport from mobile and wireless sources. The data to be transmitted is too much compared to link capacity. Labelling data and transmit only useful part of the collected data or their features is a promising solution for this challenge. Abnormal data are valuable due to the need to train models and to detect anomalies when being compared to already overflowing normal data. Labelling can be done in data sources or edges to balance the load and computing between sources, edges, and centres. However, unsupervised labelling method is still a challenge preventing to implement the above solutions. Two main problems in unsupervised labelling are long-term dynamic multiseasonality and heteroscedasticity. This paper proposes a data-driven method to handle modelling and heteroscedasticity problems. The method contains the following main steps. First, raw data are preprocessed and grouped. Second, main models are built for each group. Third, models are adapted back to the original measured data to get raw residuals. Fourth, raw residuals go through deheteroscedasticity and become normalized residuals. Finally, normalized residuals are used to conduct anomaly detection. The experimental results with real-world data show that our method successfully increases receiver-operating characteristic (AUC) by about 30%.

LONG-TERM CORRELATIONS AND MULTIFRACTALITY OF TRAFFIC FLOW MEASURED BY GIS FOR CONGESTED AND FREE-FLOW ROADS

Fractals ◽  
2016 ◽  
Vol 24 (01) ◽  
pp. 1650012 ◽  
Author(s):  
BAOFENG DI ◽  
KAI SHI ◽  
KAISHAN ZHANG ◽  
LAURENCE SVIRCHEV ◽  
XIAOXI HU
Keyword(s):  
Time Series ◽  
Real Time ◽  
Process Analysis ◽  
Moving Average ◽  
Free Flow ◽  
Traffic Information ◽  
Fluctuation Analysis ◽  
Traffic Data ◽  
Multifractal Detrended Fluctuation Analysis

In this paper, a GIS-based method was developed to extract the real-time traffic information (RTTI) from the Google Maps system for city roads. The method can be used to quantify both congested and free-flow traffic conditions. The roadway length was defined as congested length (CL) and free-flow length (FFL). Chengdu, the capital of Sichuan Province in the southwest of China, was chosen as a case study site. The RTTI data were extracted from the Google real-time maps in May 12–17, 2013 and were used to derive the CL and FFL for the study areas. The Multifractal Detrended Fluctuation Analysis (MFDFA) was used to characterize the long-term correlations of CL and FFL time series and their corresponding multifractal properties. Analysis showed that CL and FFL had demonstrated time nonlinearity and long-term correlations and both characteristics differed significantly. A shuffling procedure and a phase randomization procedure were further integrated with multifractal detrending moving average (MFDMA) to identify the major sources of multifractality of these two time series. The results showed that a multifractal process analysis could be used to characterize complex traffic data. Traffic data collected and methods developed in this paper will help better understand the complex traffic systems.

An improved fruit fly algorithm-unscented Kalman filter-echo state network method for time series prediction of the network traffic data with noises

2020 ◽  
Vol 42 (7) ◽  
pp. 1281-1293
Author(s):  
Ying Han ◽  
Yuanwei Jing ◽  
Georgi M Dimirovski
Keyword(s):  
Time Series ◽  
Kalman Filter ◽  
Network Traffic ◽  
Unscented Kalman Filter ◽  
Prediction Method ◽  
Fruit Fly ◽  
Traffic Prediction ◽  
Echo State Network ◽  
Time Varying ◽  
Traffic Data

With the complexity of the network system rapidly increasing, network traffic prediction has great significance for the safety pre-warning of the network load, network management and control, and improvement of the quality of the network service. In this paper, the time series analysis is used for the network traffic prediction, and a prediction method combined with an optimized unscented Kalman filter (UKF) by an improved fruit fly algorithm (IFOA) and echo state network (ESN) is proposed, which is named by IFOA-UKF-ESN. The researches mainly solve the problem that the prediction accuracy might be greatly affected by the actual network traffic data with unknown and time-varying noises. UKF is used to train the best state vector (formed by spectral radius, scale of the reservoir, scale of the input units and connectivity rate) of ESN; and the proposed IFOA algorithm is proposed to optimize the weights of the predicted state value and the covariance in UKF, which makes UKF have adaptive ability for unknown and time-varying noise. Three actual network traffic data sets with different Gaussian white noise distributions are constructed for experiments, and the experimental results show that the proposed prediction method makes an average improvement by reducing at least 20.60%, 43.23% and 41.85% of RMSE, at least 23.66%, 52.38% and 47.50% of MAE, and at least 23.58%, 52.10% and 47.28% of MAPE, which verify the effectiveness of the proposed method.

878-P: Real-World Data from U.S. Patients Using a Long-Term Implantable Continuous Glucose Monitoring (CGM) System: Age Effect on Glycemic Control

Diabetes ◽  
2020 ◽  
Vol 69 (Supplement 1) ◽  
pp. 878-P
Author(s):  
KATHERINE TWEDEN ◽  
SAMANWOY GHOSH-DASTIDAR ◽  
ANDREW D. DEHENNIS ◽  
FRANCINE KAUFMAN
Keyword(s):  
Glycemic Control ◽  
Continuous Glucose Monitoring ◽  
Real World ◽  
Glucose Monitoring ◽  
Age Effect ◽  
Real World Data ◽  
World Data

PREDICTING CHARACTERISTICS OF SELF SIMILAR TRAFFIC BY SPLINE-EXTRAPOLATION

2019 ◽  
Author(s):  
Irina Strelkovskay ◽  
Irina Solovskaya ◽  
Anastasija Makoganjuk ◽  
Nikolaj Severin
Keyword(s):  
Quality Characteristics ◽  
Traffic Prediction ◽  
Normative Values ◽  
Spline Functions ◽  
Network Nodes ◽  
Traffic Forecasting ◽  
Network Equipment ◽  
Self Similar ◽  
Accuracy Of Prediction

The problem of forecasting self-similar traffic, which is characterized by a considerable number of ripples and the property of long-term dependence, is considered. It is proposed to use the method of spline extrapolation using linear and cubic splines. The results of self-similar traffic prediction were obtained, which will allow to predict the necessary size of the buffer devices of the network nodes in order to avoid congestion in the network and exceed the normative values ​​of QoS quality characteristics. The solution of the problem of self-similar traffic forecasting obtained with the Simulink software package in Matlab environment is considered. A method of extrapolation based on spline functions is developed. The proposed method has several advantages over the known methods, first of all, it is sufficient ease of implementation, low resource intensity and accuracy of prediction, which can be enhanced by the use of quadratic or cubic interpolation spline functions. Using the method of spline extrapolation, the results of self-similar traffic prediction were obtained, which will allow to predict the required volume of buffer devices, thereby avoiding network congestion and exceeding the normative values ​​of QoS quality characteristics. Given that self-similar traffic is characterized by the presence of "bursts" and a long-term dependence between the moments of receipt of applications in this study, given predetermined data to improve the prediction accuracy, it is possible to use extrapolation based on wavelet functions, the so-called wavelet-extrapolation method. Based on the results of traffic forecasting, taking into account the maximum values ​​of network node traffic, you can give practical guidance on how traffic is redistributed across the network. This will balance the load of network objects and increase the efficiency of network equipment.

scholarly journals Aerosol optical depth retrievals at the Izaña Atmospheric Observatory from 1941 to 2013 by using artificial neural networks

2016 ◽  
Vol 9 (1) ◽  
pp. 53-62 ◽  
Author(s):  
R. D. García ◽  
O. E. García ◽  
E. Cuevas ◽  
V. E. Cachorro ◽  
A. Barreto ◽  
...  
Keyword(s):  
Neural Networks ◽  
Time Series ◽  
Artificial Neural Networks ◽  
Aerosol Optical Depth ◽  
Optical Depth ◽  
Mineral Dust ◽  
Dust Particles ◽  
Filter Radiometer ◽  
Artificial Neural

Abstract. This paper presents the reconstruction of a 73-year time series of the aerosol optical depth (AOD) at 500 nm at the subtropical high-mountain Izaña Atmospheric Observatory (IZO) located in Tenerife (Canary Islands, Spain). For this purpose, we have combined AOD estimates from artificial neural networks (ANNs) from 1941 to 2001 and AOD measurements directly obtained with a Precision Filter Radiometer (PFR) between 2003 and 2013. The analysis is limited to summer months (July–August–September), when the largest aerosol load is observed at IZO (Saharan mineral dust particles). The ANN AOD time series has been comprehensively validated against coincident AOD measurements performed with a solar spectrometer Mark-I (1984–2009) and AERONET (AErosol RObotic NETwork) CIMEL photometers (2004–2009) at IZO, obtaining a rather good agreement on a daily basis: Pearson coefficient, R, of 0.97 between AERONET and ANN AOD, and 0.93 between Mark-I and ANN AOD estimates. In addition, we have analysed the long-term consistency between ANN AOD time series and long-term meteorological records identifying Saharan mineral dust events at IZO (synoptical observations and local wind records). Both analyses provide consistent results, with correlations  >  85 %. Therefore, we can conclude that the reconstructed AOD time series captures well the AOD variations and dust-laden Saharan air mass outbreaks on short-term and long-term timescales and, thus, it is suitable to be used in climate analysis.

Sign in / Sign up

Export Citation Format

Share Document