Traditional Time Series Methods Research Articles

Prediction of Liaoning province steel import and export trade based on deep learning models

AbstractIn the field of deep learning, time series forecasting, particularly for economic and trade data, is a critical area of research. This study introduces a hybrid of auto regressive integrated moving average and gated recurrent unit (ARIMA‐GRU) to enhance the prediction of steel import and export trade in Liaoning Province, addressing the limitations of traditional time series methods. Traditional models like ARIMA excel with linear data but often struggle with non‐linear patterns and long‐term dependencies. The ARIMA‐GRU model combines ARIMA's linear data analysis with GRU's proficiency in non‐linear pattern recognition, effectively capturing complex dynamics in economic datasets. Our experiments show that this hybrid approach surpasses traditional models in accuracy and reliability for forecasting steel trade, providing valuable insights for economic planning and strategic decision‐making. This innovative approach not only advances the field of economic forecasting but also demonstrates the potential of integrating deep learning techniques in complex data analysis.

NDVI time-series data reconstruction for spatial-temporal dynamic monitoring of Arctic vegetation structure

ABSTRACT Spatial-temporal dynamics monitoring of Arctic vegetation structure (i.e. distribution range of tundra and forest) is of great significance for evaluating global warming effect. Currently, time-series monitoring of Arctic vegetation structure relies primarily on the Normalized Difference Vegetation Index (NDVI), which is derived from optical remote sensing images. However, because of factors such as the long revisit period of satellites and the impact of climate, optical observations are severely lacking in the Arctic region. This results in NDVI time-series data highly discontinuous and difficult to reflect actual variations in Arctic vegetation structure, and the traditional time-series reconstruction method would usually fail for severe missing conditions. Therefore, this study developed a Time Series Reconstruction method considering Periodic Trend (TSR-PT), which is specifically for alleviating the severe missing observation condition in the Arctic region. It can separate the phenological change and trend change of the incomplete time series NDVI, and borrow the information from the neighboring unchanged years for compensate of the missing observations in current years, based on the learned inter-annual and intra-annual correlation. We explore its usability in monitoring vegetation structure variation in Vorkuta region (transition zone of tundra and taiga in the Arctic Circle) based on MODIS data. It is found that the proposed TSR-PT is able to reconstruct NDVI with reasonable phenological feature even the missing rate reaches over 70%, which is usually falsely constructed by traditional filtering or fitting method, and suppress them by 0.038 in terms of RMSE; besides, we find that since 21-century, the Arctic trees have continued to increase and encroach the original tundra ecosystem, which caused a largely Arctic vegetation structural change, and we believe the proposed method would largely promote the Arctic vegetation research.

An investigation of the price discovery role of futures markets: A dynamic time warping analysis of the United States corn markets

AbstractFutures markets are critical to price discovery and often dominate spot markets. We analyze the linkages between daily corn futures and spot prices in the United States using dynamic time warping. This nonparametric pattern recognition technique has several advantages over traditional time series methods. First, it can detect multiple changes in the lead‐lag associations between the two prices within short intervals; the duration with which one series leads or lags another is not assumed to be fixed. Second, the method can be applied to time series without regard to their stationarity properties. This greatly expands the scope of this method to accommodate a wide range of time series. Third, it lends itself well to studying small samples, which econometricians encounter routinely. Fourth, the results are presented intelligibly using intuitive visualizations. Our results show that futures markets are critical to price discovery; nevertheless, spot markets dominate futures markets intermittently. We discuss the results in detail, setting them in the proper context. [EconLit Citations: C14, C32, Q02, Q11].

Spatial-temporal load forecasting of electric vehicle charging stations based on graph neural network

Accurate forecasting of the load of electric vehicle (EV) charging stations is critical for EV users to choose the optimal charging stations and ensure the safe and efficient operation of the power grid. The charging load of different charging stations in the same area is interrelated. However, forecasting the charging load of individual charging station using traditional time series methods is insufficient. To fully consider the spatial-temporal correlation between charging stations, this paper proposes a new charging load forecasting framework based on the Adaptive Spatial-temporal Graph Neural Network with Transformer (ASTNet-T). First, an adaptive graph is constructed based on the spatial relationship and historical information between charging stations, and the local spatial-temporal dependencies hidden therein are captured by the spatio-temporal convolutional network. Then, a Transformer network is introduced to capture the global spatial-temporal dependencies of charging loads and predict the future multilevel charging loads of charging stations. Finally, extensive experiments are conducted on two real-world charging load datasets. The effectiveness and robustness of the proposed algorithm are verified by experiments. In the Dundee City dataset, the MAE, MAPE, and RMSE values of the proposed model are improved by approximately 71%, 90%, and 67%, respectively, compared to the suboptimal baseline model, demonstrating that the proposed algorithm significantly improves the accuracy of load forecasting.

Analyzing the Effectiveness of Ensemble Based Analysis in Wireless Sensor Networks

The usefulness of ensemble-based total time series analysis in Wi-Fi sensor networks is examined in this paper. A device to uses an ensemble approach combines multiple strategies to enhance overall predictive performance. This research assesses various tactics using unique metrics, such as robustness and accuracy. It contrasts the effectiveness of traditional time series methods with ensemble-based total fashions. An experimental approach focusing mostly on exceptional Wi-Fi sensor network scenarios is employed to evaluate the overall effectiveness of the suggested methods. Additionally, this study looks into how changes to community features like energy delivery, conversation range, and node density affect how effective the suggested methods are. The study's findings maintain the capacity to create effective Wi-Fi sensor networks with improved predicted overall performance. The usefulness of ensemble-based time collecting and analysis techniques for wireless sensor networks is investigated in this research. This study primarily looks upon function extraction and seasonality discounting of time series records in WSNs. In this analysis, seasonality is discounted using an ensemble median filter, and feature extraction is accomplished by primary component assessment. To assess the performance of the suggested ensemble technique on every simulated and real-world international WSN fact, multiple experiments are carried out. The findings suggest that the ensemble approach can improve the exceptional time-gathering records within WSNs and reduce seasonality. Furthermore, when compared to single-sensor strategies, the ensemble technique further improves the accuracy of the function extraction system. This work demonstrates the applicability of the ensemble approach for the investigation of time collection data in WSNs

CO2 concentration forecasting in smart cities using a hybrid ARIMA–TFT model on multivariate time series IoT data

Carbon Dioxide (CO_{2}) is a significant contributor to greenhouse gas emissions and one of the main drivers behind global warming and climate change. In spite of the global economic slowdown due to the COVID-19 pandemic, the global average atmospheric CO_{2} concentration reached a new record high in 2020 with its year-on-year increase being the fifth highest annual increase in 63 years, according to the National Oceanic and Atmospheric Administration. Furthermore, the years 2020 and 2019 were respectively the second and third warmest, while the decade 2010–2019 was the warmest decade ever recorded. In an attempt to curb this climate emergency, many countries and organizations globally have adopted ambitious goals and announced plans to help dramatically reduce CO_{2} emissions. As part of these plans, various innovative smart city projects are being developed, focusing on implementing Internet of Things (IoT) technologies. By collecting sensor-based data, such technologies aim towards automating data-driven decision-making around carbon emission management and reduction. In this work, a hybrid machine learning system, aimed at forecasting CO_2 concentration levels in a smart city environment was developed using a multivariate time series dataset containing IoT sensor measurements of CO_{2}, as well as various environmental factors, taken at every second. The proposed system demonstrated superior performance to similar methods, while also maintaining a high degree of interpretability. More specifically, the approach was empirically compared against other similar approaches in several scenarios and use cases, thus also offering more insight into the predictive capabilities of such state-of-the-art systems. For this comparison, both traditional time series and deep learning approaches were employed, including the current state-of-the-art architectures, such as attention-based, transformer networks. Results demonstrated that, when measured across various settings and metrics, including three different forecasting horizons, the hybrid solution achieved the best overall results, and in some cases, the difference in performance was statistically significant. At the same time, insights from the system’s inner workings were extracted, shedding light on the reasoning behind the model’s predictions and the factors that contribute to them, thus showcasing its transparency. Lastly, throughout the experiments, deep learning approaches illustrated their ability to better handle the multivariate nature of the dataset and in general tended to outperform the traditional time series methods, especially for longer forecasting horizons.

Local Projections for Applied Economics

The dynamic causal effect of an intervention on an outcome is of paramount interest to applied macro- and microeconomics research. However, this question has been generally approached differently by the two literatures. In making the transition from traditional time series methods to applied microeconometrics, local projections can serve as a natural bridge. Local projections can translate the familiar language of vector autoregressions and impulse responses into the language of potential outcomes and treatment effects. There are gains to be made by both literatures from greater integration of well-established methods in each. This review shows how to make these connections and points to potential areas of further research.

A MACHINE LEARNING APPROACH TO GDP NOWCASTING:AN EMERGING MARKET EXPERIENCE

The growth rate of real Gross Domestic Product (GDP), as measured by the National Statistical Office of India, is an important metric for monetary policy making. Because GDP is released with a significant lag, particularly for the emerging market economies, this article presents various methodologies for nowcasting and forecasting GDP, using both traditional time series and machine learning methods. Further, considering the importance of forward-looking information, our nowcasting model incorporates financial market data and an economic uncertainty index, in addition to high-frequency traditional macroeconomic indicators. Our findings suggest an improvement in the performance of nowcasting using a hybrid of machine learning and conventional time series methods.

Single-variable method for predicting trends in chlorophyll a concentration based on the similarity of time series

Algal blooms are increasing in global water bodies, which threatens social development. Predicting the chlorophyll a (Chla) concentration is useful for water environmental management. Most previous studies of Chla prediction required multiple variables and considerable computation. Therefore, a similarity-based prediction algorithm (SPA) was developed for this prediction. The mean relative error (MRE) and dynamic time warping (DTW) algorithms are similarity measures used to create SPAs (SPA-MRE, SPA-DTW and SPA-MRE&DTW).Model validation was performed at the Yuqiao Reservoir north of Tianjin, China. The Chla concentration dataset of the Yuqiao Reservoir contains hourly data for 745 days from 2017 to 2020. According to the model validation results, SPA methods performed better than traditional time-series analysis methods (ARIMA (−9.7%) and Holt-Winters (−21.2%)). The MRE predicted by the SPA method at 24 h was 0.187.These three SPA methods showed different characteristics and applicability. SPA-MRE was suitable for predicting low Chla concentrations. SPA-DTW was suitable for predicting high Chla concentrations with obvious fluctuations. SPA-MRE combined with DTW exhibited the best comprehensive performance. Moreover, the SPA maintained high robustness and reliability even in the absence of 9% of the data.SPA is a single-variable prediction method that is more cost-effective than traditional regression-based and mechanism-based methods. SPA can be used with adjacent data for forecasting as in traditional time-series analysis models, and time series similar to the current series can be retrieved from historical data and used for predictions based on similarity. This study provides a robust and reliable prediction method for algal bloom management. SPA can be used to deeply mine characteristic information from historical data.

A Critical Review of Short-Term Water Demand Forecasting Tools—What Method Should I Use?

The challenge for city authorities goes beyond managing growing cities, since as cities develop, their exposure to climate change effects also increases. In this scenario, urban water supply is under unprecedented pressure, and the sustainable management of the water demand, in terms of practices including economic, social, environmental, production, and other fields, is becoming a must for utility managers and policy makers. To help tackle these challenges, this paper presents a well-timed review of predictive methods for short-term water demand. For this purpose, over 100 articles were selected from the articles published in water demand forecasting from 2010 to 2021 and classified upon the methods they use. In principle, the results show that traditional time series methods and artificial neural networks are among the most widely used methods in the literature, used in 25% and 20% of the articles in this review. However, the ultimate goal of the current work goes further, providing a comprehensive guideline for engineers and practitioners on selecting a forecasting method to use among the plethora of available options. The overall document results in an innovative reference tool, ready to support demand-informed decision making for disruptive technologies such as those coming from the Internet of Things and cyber–physical systems, as well as from the use of digital twin models of water infrastructure. On top of this, this paper includes a thorough review of how sustainable management objectives have evolved in a new era of technological developments, transforming data acquisition and treatment.

Spotlight Personnel: How Hiring and Turnover Drive Service Performance Versus Demand

In many sectors of the entertainment industry, a few employees attract the public spotlight when performing the core service. For example, in professional team sports, a team of players competes in games, and in TV shows, a cast of artists acts in different episodes. These employees, coined “spotlight personnel,” are an essential but expensive element of ongoing service delivery. Despite their importance and cost, very little is known about how changes in spotlight personnel affect service performance and demand. To address this gap, this article uses unique data on professional German soccer teams, tracking the quantity (number of players) and quality (average transfer price) of spotlight personnel hiring (incoming transfers) and turnover (outgoing transfers), objective service performance (winning percentage), and demand (ticket sales) across four decades, using both traditional and novel time series methods. The results show that service performance and demand are primarily affected by spotlight personnel hiring rather than by turnover. Hiring quantity decreases service performance yet increases demand, whereas hiring quality benefits both service performance and demand. The analysis further uncovers that these effects are subject to dynamic interactions and nonlinearities. Investment scenarios showcase how understanding these effects can substantially improve managerial decision making.

How to Handle Data Imbalance and Feature Selection Problems in CNN-Based Stock Price Forecasting

Stock market forecasting is a time series problem that aims to predict possible future prices or directions of an index/stock. The stock data contains high uncertainty and is influenced by too many factors; hence it isn’t easy to achieve the goal by traditional time series methods. In literature, the convolutional neural networks (CNN) models were used for stock market forecasting and gave successful results. But, data imbalance due to labeling and feature selection problems were seen when considering these models. Hence, this study proposed a new rule-based labeling algorithm and a new feature selection approach to solve the issues. In addition, a CNN-based model, which was presented to predict the next day’s trade action of stocks in the Dow30 index, was constructed to check the effectiveness of the data labeling and the feature selection approach. Different image-based input variable sets were created using technical indicators, gold, and oil price data to feed the CNN model. The prediction performance of CNN models was compared with other studies in the literature. The experimental results showed that the CNN prediction model, which uses the proposed feature selection and labeling approaches in this study, performs 3-22% higher accuracy than the CNN-based models taking part in other studies. Also, the labeling approach proposed is more successful than Chen and Huang’s data weighting approach to solve the stock data imbalance problem. This algorithm reduced the ratio between labeled data from 15 times to 1.8 times.

A flexible framework for intervention analysis applied to credit-card usage during the coronavirus pandemic

We develop a variant of intervention analysis designed to measure a change in the law of motion for the distribution of individuals in a cross-section, rather than modeling the moments of the distribution. To calculate a counterfactual forecast, we discretize the distribution and employ a Markov model in which the transition probabilities are modeled as a multinomial logit distribution. Our approach is scalable and is designed to be applied to micro-level data. A wide panel often carries with it several imperfections that complicate the analysis when using traditional time-series methods; our framework accommodates these imperfections. The result is a framework rich enough to detect intervention effects that not only shift the mean, but also those that shift higher moments, while leaving lower moments unchanged. We apply this framework to document the changes in credit usage of consumers during the COVID-19 pandemic. We consider multinomial logit models of the dependence of credit-card balances, with categorical variables representing monthly seasonality, homeownership status, and credit scores. We find that, relative to our forecasts, consumers have greatly reduced their use of credit. This result holds for homeowners and renters as well as consumers with both high and low credit scores.

Concurrent neural network: a model of competition between times series

Competition between times series often arises in sales prediction, when similar products are on sale on a marketplace. This article provides a model of the presence of cannibalization between times series. This model creates a "competitiveness" function that depends on external features such as price and margin. It also provides a theoretical guaranty on the error of the model under some reasonable conditions, and implement this model using a neural network to compute this competitiveness function. This implementation outperforms other traditional time series methods and classical neural networks for market share prediction on a real-world data set. Moreover, it allows controlling underprediction, which plagues traditional forecasts models.

Stock Price Prediction using Fractional Gradient-Based Long Short Term Memory

Deep Learning is considered one of the most effective strategies used by hedge funds to maximize profits. But Deep Neural Networks (DNN) lack theoretical analysis of memory exploitation. Some traditional time series methods such as Auto-Regressive Integrated Moving Average (ARIMA) and Generalized Auto-Regressive Conditional Heteroskedasticity (GARCH) work only when the entire series is pre-processed or when the whole data is available. Thus, it fails in a live trading system. So, there is a great need to develop techniques that give more accurate stock/index predictions. This study has exploited fractional-order derivatives’ memory property in the backpropagation of LSTM for stock predictions. As the history of previous stock prices plays a significant role in deciding the future price, fractional-order derivatives carry the past information along with itself. So, the use of Fractional-order derivatives with neural networks for this time series prediction is meaningful and helpful.

Visibility-graphlet approach to the output series of a Hodgkin–Huxley neuron

The output signals of neurons that are exposed to external stimuli are of great importance for brain functionality. Traditional time-series analysis methods have provided encouraging results; however, the associated patterns and their correlations in the output signals of neurons are masked by statistical procedures. Here, graphlets are employed to extract the local temporal patterns and the transitions between them from the output signals when neurons are exposed to external stimuli with selected stimulating periods. A transition network is defined where the node is the graphlet and the direct link is the transition between two successive graphlets. The transition-network structure is affected by the simulating periods. When the stimulating period moves close to an integer multiple of the neuronal intrinsic period, only the backbone or core survives, while the other linkages disappear. Interestingly, the size of the backbone (number of nodes) equals the multiple. The transition-network structure is conservative within each stimulating region, which is defined as the range between two successive integer multiples. Nevertheless, the backbone or detailed structure is significantly altered between different stimulating regions. This alternation is induced primarily from a total of 12 active linkages. Hence, the transition network shows the structure of cross correlations in the output time-series for a single neuron.

Exchange Market Liquidity Prediction with the K-Nearest Neighbor Approach: Crypto vs. Fiat Currencies

In this paper, we compare the predictions on the market liquidity in crypto and fiat currencies between two traditional time series methods, the autoregressive moving average (ARMA) and the generalized autoregressive conditional heteroskedasticity (GARCH), and the machine learning algorithm called the k-nearest neighbor (KNN) approach. We measure market liquidity as the log rates of bid-ask spreads in a sample of three cryptocurrencies (Bitcoin, Ethereum, and Ripple) and 16 major fiat currencies from 9 February 2018 to 8 February 2019. We find that the KNN approach is better suited for capturing the market liquidity in a cryptocurrency in the short-term than the ARMA and GARCH models maybe due to the complexity of the microstructure of the market. Considering traditional time series models, we find that ARMA models perform well when estimating the liquidity of fiat currencies in developed markets, whereas GARCH models do the same for fiat currencies in emerging markets. Nevertheless, our results show that the KNN approach can better predict the log rates of the bid-ask spreads of crypto and fiat currencies than ARMA and GARCH models.

Modelling the electrical energy profile of a batch manufacturing pharmaceutical facility

Sustainable manufacturing practices are a dominating consideration for legacy factories. Major attention is being applied to improving current practices to more sustainable ones. This research provides a case study of a batch manufacturing pharmaceutical facility and compares a number of approaches to modelling the electrical energy consumption in the plant. An accurate model of the electrical energy in the facility will allow more sustainable approaches to be developed. This can be achieved by improving current processes to reduce the electrical load. Historical electrical energy data were modelled using traditional time series methods. Historical manufacturing data and the electrical energy data were used to develop machine learning models using a feedforward neural network and a random forest. All of the approaches were then compared. The major challenge posed in model development and validation was acquiring data suitable for machine learning. The manufacturing data were stored in hand-written records. These records needed to be digitised and then go through a number of transformative steps before the data were suitable for modelling. The random forest model successfully modelled the energy profile of the facility. The model can be used to predict and better manage the plant electrical energy load.

Research on trend analysis method of multi-series economic data based on correlation enhancement of deep learning

The analysis on economic data based on time series takes an important position in the field of analysis on time-series data and is also an important task of the field of big data and artificial intelligence. Traditional time-series analysis method is of relatively weak competence in dealing with multi-series analysis. In this research, based on the problem associated with the analysis on time-series economic data, efficient handling method and model are put forward in the face of multi-series analysis task. Also, combined with the association rules, trend correlation and self-trend correlation among multiple series, a trend and correlation deep neural network model (TC-DNM) is established and then tested and verified by using three kinds of economic datasets with representativeness based on the trend analysis task handed by multi-series analysis. The results show that the model proposed in this research is effective than a number of baseline models, can be employed to achieve precision–recall balance and also possesses strong reusability. The two correlation models and joint models in this paper are of peculiarity and innovativeness.

Multi-Field Interference Simultaneously Imaging on Single Image for Dynamic Surface Measurement.

To obtain the dynamic surface of high-frequency vibrating mirrors (VMs), a novel method involving multi-field interference (MFI) pattern imaging on a single image is proposed in this paper. Using multiple reflections and refractions, the proposed method generates three interference patterns at the same time, which improves the traditional time-series methods where a single interference pattern can be obtained at one time. Experimental results show that a series of MFI patterns can be obtained on a single image, with the laser repetition frequency (LRF) ranging from 200 Hz to 10 Hz, and the frame rate of the camera at 10 Hz. Particularly if the LRF (10 Hz) is equal to the frame rate of image, crosstalk is avoided completely, which is particularly desirable in dynamic surface measurement. In summary, the MFI imaging method provides an effective way for VM dynamic surface measurement.