Terms Of MAPE Research Articles

GRU combined model based on multi-objective optimization for short-term residential load forecasting

Reliable and accurate short-term forecasting of residential load plays an important role in DSM. However, the high uncertainty inherent in single-user loads makes them difficult to forecast accurately. Various traditional methods have been used to address the problem of residential load forecasting. A single load forecast model in the traditional method does not allow for comprehensive learning of data characteristics for residential loads, and utilizing RNNs faces the problem of long-term memory with vanishing or exploding gradients in backpropagation. Therefore, a gated GRU combined model based on multi-objective optimization is proposed to improve the short-term residential load forecasting accuracy in this paper. In order to demonstrate the effectiveness, GRUCC-MOP is first experimentally tested with the unimproved model to verify the model performance and forecasting effectiveness. Secondly the method is evaluated experimentally with other excellent forecasting methods: models such as DBN, LSTM, GRU, EMD-DBN and EMD-MODBN. By comparing simulation experiments, the proposed GRU combined model can get better results in terms of MAPE on January, April, July, and November load data, so this proposed method has better performance than other research methods in short-term residential load forecasting.

Stock Price Prediction Using Candlestick Patterns and Sparrow Search Algorithm

Accurately forecasting the trajectory of stock prices holds crucial significance for investors in mitigating investment risks and making informed decisions. Candlestick charts visually depict price information and the trends in stocks, harboring valuable insights for predicting stock price movements. Therefore, the challenge lies in efficiently harnessing candlestick patterns to forecast stock prices. Furthermore, the selection of hyperparameters in network models has a profound impact on the forecasting outcomes. Building upon this foundation, we propose a stock price prediction model SSA-CPBiGRU that integrates candlestick patterns and a sparrow search algorithm (SSA). The incorporation of candlestick patterns endows the input data with structural characteristics and time series relationships. Moreover, the hyperparameters of the CPBiGRU model are optimized using an SSA. Subsequently, the optimized hyperparameters are employed within the network model to conduct predictions. We selected six stocks from different industries in the Chinese stock market for experimentation. The experimental results demonstrate that the model proposed in this paper can effectively enhance the prediction accuracy and has universal applicability. In comparison to the LSTM model, the proposed model produces an average of 31.13%, 24.92%, and 30.42% less test loss in terms of MAPE, RMSE and MAE, respectively. Moreover, it achieves an average improvement of 2.05% in R2.

EV charging load forecasting model mining algorithm based on hybrid intelligence

Precise EV charging load forecasting plays a critical role in optimizing resource allocation and facilitating economic operation and energy management of EV charging stations, as well as supporting the economic dispatch of the power grid for efficient and effective utilization of resources. Existing EV charging load forecasting models have highly restrictive load data requirements and thus have practical limitations in two fold: first, these models are a black box and cannot provide a quantitative reference for later analysis affecting EV charging load forecasting; and second, they fail in considering the impact of noisy data caused by uncontrollable factors (data collection failures, human errors, and network attacks, etc.) on the accuracy of EV charging load forecasting. To address the above issues, we propose a novel EV charging load forecasting model mining based on gene expression programming (CFMM-GEP) by fusing noisy load processing. This will tackle three-fold ideas: (1) the charging load dataset with abnormal data is reconstructed based on the Auto-Encoder, and the support vector machine-based abnormal detection algorithm is proposed; (2) a quantitative model for EV charging load forecasting based on gene expression programming is constructed. Extensive experiments are carried out on four open-source charging load datasets. Experimental results indicate that the superiority of our proposed CFMM-GEP model over 6 state-of-the-art models in terms of MAPE, RMSE, MAE, and R2.

CA-LSTM: An Improved LSTM Trajectory Prediction Method Based on Infrared UAV Target Detection

In order to improve the UAV prevention and control capability in key areas and improve the rapid identification and trajectory prediction accuracy of the ground detection system in anti-UAV early warnings, an improved LSTM trajectory prediction network CA-LSTM (CNN-Attention-LSTM) based on attention enhancement and convolution fusion structure is proposed. Firstly, the native Yolov5 network is improved to enhance its detection ability for small targets of infrared UAVs, and the trajectory of UAVs in image space is constructed. Secondly, the LSTM network and convolutional neural network are integrated to improve the expression ability of the deep features of UAV trajectory data, and at the same time, the attention structure is designed to more comprehensively obtain time series context information, improve the influence on important time series features, and realize coarse–fine-grained feature fusion. Finally, tests were carried out on a homemade UAV infrared detection dataset. The experimental results show that the algorithm proposed in this paper can quickly and accurately identify infrared UAV targets and can achieve more accurate predictions of UAV flight trajectories, which are reduced by 9.43% and 23.81% in terms of MAPE and MSE indicators compared with the native LSTM network (the smaller the values of these evaluation indexes, the better the prediction results).

Optimal fusion‐based localization method for tracking of smartphone user in tall complex buildings

AbstractIn the event of a fire breaking out or in other complicated situations, a mobile computing solution combining the Internet of Things and wearable devices can actually assist tracking solutions for rescuing and evacuating people in multistory structures. Thus, it is crucial to increase the positioning technology's accuracy. The sequential Monte Carlo (SMC) approach is used in various applications such as target tracking and intelligent surveillance, which rely on smartphone‐based inertial data sequences. However, the SMC method has intrinsic flaws, such as sample impoverishment and particle degeneracy. A novel SMC approach is presented, which is built on the weighted differential evolution (WDE) algorithm. Sequential Monte Carlo approaches start with random particle placements and arrives at the desired distribution with a slower variance reduction, like in a high‐dimensional space, such as a multistory structure. Weighted differential evolution is included before the resampling procedure to guarantee the appropriate variety of the particle set, prevent the usage of an inadequate number of valid samples, and preserve smartphone user position accuracy. The values of the smartphone‐based sensors and BLE‐beacons are set as input to the SMC, which aids in fast approximating the posterior distributions, to speed up the particle congregation process in the proposed SMC‐based WDE approach. Lastly, the robustness and efficacy of the suggested technique more accurately reflect the actual situation of smartphone users. According to simulation findings, the suggested approach provides improved location estimation with reduced localization error and quick convergence. The results confirm that the proposed optimal fusion‐based SMC‐WDE scheme performs 9.92% better in terms of MAPE, 15.24% for the case of MAE, and 0.031% when evaluating based on the R2 Score.

Development of a Short-Term Electrical Load Forecasting in Disaggregated Levels Using a Hybrid Modified Fuzzy-ARTMAP Strategy

In recent years, electrical systems have evolved, creating uncertainties in short-term economic dispatch programming due to demand fluctuations from self-generating companies. This paper proposes a flexible Machine Learning (ML) approach to address electrical load forecasting at various levels of disaggregation in the Peruvian Interconnected Electrical System (SEIN). The novelty of this approach includes utilizing meteorological data for training, employing an adaptable methodology with easily modifiable internal parameters, achieving low computational cost, and demonstrating high performance in terms of MAPE. The methodology combines modified Fuzzy ARTMAP Neural Network (FAMM) and hybrid Support Vector Machine FAMM (SVMFAMM) methods in a parallel process, using data decomposition through the Wavelet filter db20. Experimental results show that the proposed approach outperforms state-of-the-art models in predicting accuracy across different time intervals.

GlottisNetV2: Temporal Glottal Midline Detection Using Deep Convolutional Neural Networks.

High-speed videoendoscopy is a major tool for quantitative laryngology. Glottis segmentation and glottal midline detection are crucial for computing vocal fold-specific, quantitative parameters. However, fully automated solutions show limited clinical applicability. Especially unbiased glottal midline detection remains a challenging problem. We developed a multitask deep neural network for glottis segmentation and glottal midline detection. We used techniques from pose estimation to estimate the anterior and posterior points in endoscopy images. Neural networks were set up in TensorFlow/Keras and trained and evaluated with the BAGLS dataset. We found that a dual decoder deep neural network termed GlottisNetV2 outperforms the previously proposed GlottisNet in terms of MAPE on the test dataset (1.85% to 6.3%) while converging faster. Using various hyperparameter tunings, we allow fast and directed training. Using temporal variant data on an additional data set designed for this task, we can improve the median prediction accuracy from 2.1% to 1.76% when using 12 consecutive frames and additional temporal filtering. We found that temporal glottal midline detection using a dual decoder architecture together with keypoint estimation allows accurate midline prediction. We show that our proposed architecture allows stable and reliable glottal midline predictions ready for clinical use and analysis of symmetry measures.

Air quality indicators and AQI prediction coupling long-short term memory (LSTM) and sparrow search algorithm (SSA): A case study of Shanghai

Air quality indicators and air quality index (AQI) prediction are effective approaches for urban decision-makers, planners, managers and even city residents to arrange their risk abatement measures in advance. In this study, five models, including Random Forest (RF), eXtreme Gradient Boosting (XGBoost), Multi-layer perceptron (MLP), Long-short term memory (LSTM) and Long-short term memory coupling with sparrow search algorithm (LSTM-SSA), were used to predict hourly, daily and weekly concentrations of six air quality indicators (PM2.5, PM10, SO2, NO2, CO, O3) and AQI. The case study was based on the hourly observed data of Shanghai from February 1, 2021 to January 31, 2022 and prediction accuracy for different prediction models, indicators and periods was compared. Results revealed that: (1) The prediction accuracy of three neural network models (MLP, LSTM, LSTM-SSA) was superior to two tree-based models (RF, XGBoost) in terms of MAPE, RMSE, MAE and R2. (2) Comparing three neural network models, LSTM-SSA was more accurate than MLP and LSTM in terms of MAPE, RMSE, MAE and R2. (3) For LSTM-SSA, the MAPE of AQI was minimal (1.53%), followed by PM2.5 (2.72%), O3 (3.93%), PM10 (4.87%), NO2 (5.74%), CO (7.54%) and SO2 (8.82%). (4) For LSTM-SSA, the value of MAPE increased from 3.88%, 5.28%–5.91% when the prediction period increased from an hour, a day to a week.

Convolutional neural network forecasting of European Union allowances futures using a novel unconstrained transformation method

This paper develops an open-high-low-close (OHLC) data forecasting framework to forecast EUA futures price based on EU ETS data and extended exogenous variables from 2013 to 2020. The challenge of forecasting such an OHLC structure lies in handling its three intrinsic constraints, i.e., the positive constraint, interval constraint, and boundary constraint. This paper proposes a novel unconstrained transformation method for OHLC data and combines it with various forecasting models. Out-of-sample modelings identify the extraordinary performance of the convolutional neural network (CNN) in terms of MAPE (1.371%), MAE (0.274), RMSE (0.370), and AR (0.621), better than that of multiple linear regression (MLR), vector auto-regression (VAR) and vector error correction model (VECM), support vector regression (SVR), and multi-layer perceptron (MLP). The proposed transformation-based forecasting framework demonstrates the considerable potential for OHLC data forecasting in the energy finance field, e.g., crude and natural gas. Practicable and concrete suggestions are provided to ensure the profitability of trading EUA futures.

A New Hybrid Model of Fuzzy Time Series and Genetic Algorithm Based Machine Learning Algorithm: A Case Study of Forecasting Prices of Nine Types of Major Cryptocurrencies

This study pertains to the usage and effectiveness of the fuzzy time series (FTS) models and machine learning methods in forecasting movements of financial data. The datasets used in this study are the actual closing prices and transaction volumes of 9 different cryptocurrencies, from the earliest time obtainable in Yahoo! Finance, all the way to 31 Oct 2021. Firstly, this paper presents a study of the severe drawbacks of all existing literature on FTS. In particular, this article outlines severe shortcomings of all existing FTS based algorithms that caused inaccuracies among all existing FTS-based algorithms in yielding meaningful prediction. Then, a novel structure of our improvised FTS, denoted as QFTS, is presented in this paper, which aims to rectify all flaws exist in all conventional FTS based models in literature. A further hybrid of QFTS with ANN is also presented. Later, a comparative analysis of all the aforementioned FTS models is presented in terms of overall forecasting accuracy and forecasting accuracy under specific conditions. The results are being compared in terms of MAPE. The newly invented QFTS model and the QFTS-ANN hybrid is found to profoundly outperform all the existing FTS models in literature, which includes Singh's FTS model. Such innovation profoundly rectifies severe shortcomings in financial forecasting that have persisted for many years in the past literature.

Spatiotemporal attention mechanism-based multistep traffic volume prediction model for highway toll stations

As the fundamental part of other Intelligent Transportation Systems (ITS) applications, short-term traffic volume prediction plays an important role in various intelligent transportation tasks, such as traffic management, traffic signal control and route planning. Although Neural-network-based traffic prediction methods can produce good results, most of the models can’t be explained in an intuitive way. In this paper, we not only proposed a model that increase the short-term prediction accuracy of the traffic volume, but also improved the interpretability of the model by analyzing the internal attention score learnt by the model. we propose a spatiotemporal attention mechanism-based multistep traffic volume prediction model (SAMM). Inside the model, an LSTM-based Encoder-Decoder network with a hybrid attention mechanism is introduced, which consists of spatial attention and temporal attention. In the first level, the local and global spatial attention mechanisms considering the micro traffic evolution and macro pattern similarity, respectively, are applied to capture and amplify the features from the highly correlated entrance stations. In the second level, a temporal attention mechanism is employed to amplify the features from the time steps captured as contributing more to the future exit volume. Considering the time-dependent characteristics and the continuity of the recent evolutionary traffic volume trend, the timestamp features and historical exit volume series of target stations are included as the external inputs. An experiment is conducted using data from the highway toll collection system of Guangdong Province, China. By extracting and analyzing the weights of the spatial and temporal attention layers, the contributions of the intermediate parameters are revealed and explained with knowledge acquired by historical statistics. The results show that the proposed model outperforms the state-of-the-art model by 29.51% in terms of MSE, 13.93% in terms of MAE, and 5.69% in terms of MAPE. The effectiveness of the Encoder-Decoder framework and the attention mechanism are also verified.

Closed-Loop Aggregated Baseline Load Estimation Using Contextual Bandit With Policy Gradient

Demand response (DR) is an important technique to explore the demand-side flexibility. The wide deployment of smart meters makes it possible to quantify the baseline load. As an intermediate agent, demand response aggregator needs to obtain the aggregated baseline load (ABL) for the DR event. Previous studies about the household level estimation focus on the estimation method. However, for ABL estimation, customer division is an important issue. A major limitation is the mismatch between the objectives of segmentation and estimation. Therefore, this paper proposes a new closed-loop method for estimating the ABL, which utilizes the contextual bandit with policy gradient to link the segmentation with the estimation. As such, the ABL estimation accuracy can guide the segmentation to divide the customers. The segmentation and estimation optimize collaboratively to improve the ABL estimation accuracy. An ensemble method for combining network’s weights during the training process is proposed. Moreover, a pre- and post-event adjustment method is developed to further improve the estimation accuracy. Comprehensive comparisons demonstrate the proposed method can achieve the best estimation performance with regard to the MAPE and RMSE. It improves the estimation accuracy by 7% in terms of MAPE, and 11% in terms of RMSE.

Development of a Cloud-Based Real-Time Building Health Monitoring and Prediction System Using AI and IoT Sensors

In this study, a cloud-based real-time building health monitoring and prediction system using AI and IoT sensors was developed. To predict the building condition, which constitutes time-series data, statistical-based ARIMA and AI-based LSTM prediction models were designed, and the effectiveness of the proposed prediction models was experimentally verified using a 1/8-scaled miniaturized structure. The prediction accuracy in terms of MAPE (less than 1%) was experimentally confirmed to be satisfactory. Moreover, a method for analyzing dimensional structure deformation was developed by combining multiple sensor measurements, and its effectiveness was verified through the case study of a real earthquake-damaged building.

Distance measures in building informatics: An in-depth assessment through typical tasks in building energy management

Distance measurement (also known as similarity measurement) is used to evaluate pairwise similarities between data samples. It has been widely used in diverse building informatics research and applications to classify or cluster massive building data with the aim of improving prediction accuracy, identifying operation patterns, benchmarking and diagnosing building performance, etc. Various distance measures have been adopted to measure the distance/similarity of building data. However, the intrinsic complexity and diversity of building operational data bring considerable difficulties to the selection of a suitable distance measure for a specific task. There is a strong and urgent need for a comprehensive review and systematic comparison of existing distance measures in building informatics.This study provides a comprehensive review of various distance measures and their applications in building operational data analysis. A systematic comparison is undertaken based on two typical tasks relying on building informatics, i.e., building energy usage pattern recognition, and clustering-based weather data segmentation for the customized development of building energy prediction models. Nine widely adopted distance measures have been reviewed and compared, including Euclidean distance, Chebyshev distance, Manhattan distance, Mahalanobis distance, Hausdorff distance, Pearson correlation distance, Dynamic Time Warping, Edit distance on Real Sequence, and Cosine distance. Novel internal and external clustering validation approaches based on the cross-test and prediction accuracy are proposed and adopted to compare the clustering performance. The results in case studies showed that weather data clustering using the Cosine distance and Pearson correlation distance helps to obtain better energy prediction results in terms of MAPE (13.22% and 12.91%, respectively) than the commonly-used Euclidean distance (13.99%). The results also revealed that better clustering performance does not necessarily lead to higher prediction accuracy. The research results and insights obtained are valuable to guide distance-based research in building informatics.

Prediction and Analysis of Short-Term Load Forecasting Model Based on Similar Day Clustering and CatBoost

Accurate load forecasting provides a reference of vital importance for power generation and power dispatching systems. A short-term load forecasting model combining similar day clustering and CatBoost method is proposed in this paper. Similar day clustering is a method for evaluating the similarity between the days to be forecast and historical days. Similar day clustering is based on K-means clustering which can classify samples with similar features into the same category. Then, CatBoost is used to build load forecasting models. CatBoost enables us to directly handle categorical features without pre-processing. This paper utilized the information collected by a substation in Hangzhou as the data set, and compare the proposed model with the existing forecasting model based on Autoregressive Moving Average, Gradient Boosting Decision Tree, CatBoost, Long Short-Term Memory. The experimental results show that the proposed model provides more accurate results compared with other models in terms of MAPE, RMSE.

Short-term prediction of airway congestion index using machine learning methods

This study presents a hybrid spatiotemporal convolutional long short-term memory network with tri-directional temporal features (SCLN-TTF) for airway congestion index prediction. The spatial features, temporal features and spatiotemporal features are taken as input variables, among which the temporal and spatiotemporal input datasets are categorized into forward time series datasets, backward time series datasets and datasets at the target time. The spatial features are extracted using the fast-approximate convolutions on graphs, and the temporal features are extracted using the long short-term memory network. The subsequent congestion index is the output variable. The model parameters are sequentially updated based on the recent collected data and the new predicting results. It is found that, in general, the proposed SCLN-TTF method achieves the most stable and satisfactory performance for all the selected airways. The models exhibit the best performance for the 60 min prediction time horizon in terms of MAPE, while the proposed SCLN-TTF shows superiority compared to the other benchmark methods particularly for the 10 min prediction. Then the graph convolutional network and extreme gradient boosting method are combined for high congestion index identification. The finding further confirms the superiority of the proposed method by considering both the spatial and temporal dependencies and integrating forward and backward features.

Investigating data-driven approaches to understand the interaction between water quality and physiological response of sentinel oysters in natural environment

The research presented in this paper was conducted as part of a project that aimed at using biosensors on sentinel oysters to provide a biological perspective of environmental changes. The physiological response patterns (e.g. heart rate variability, shell gape) of sentinel oysters can provide valuable insight into their exposure to environmental stressors. Sensors were attached to measure both heart rate and shell gape (physiological behaviour) of oysters placed at different depths in the water column. Sensors were also deployed to measure water quality that represents the environmental condition the oysters were in. The objective of this study is to utilise the data from different sensors to investigate how environmental conditions modulate physiological response of oysters. We have utilised a set of machine learning models to develop data-driven approaches that can predict heart rate and shell gape (opening and closing action) of oysters from water quality variables, and vice-versa. The level of prediction accuracy indicates how well environmental conditions influence the physiological response of oysters. The effectiveness of the developed approaches is evaluated using data collected from two different deployments of sensors in South East Tasmania, Australia. Experimental results demonstrate that the presented data-driven approaches can provide accurate predictions of physiological and water quality variables, for the data set considered in this study. The prediction error (in terms of MAPE) for HR and water quality is in the range of 3.17%-8.21% and 0.72%–2.48%, respectively, and classification accuracy (F-Score) for shell gape varies between 0.96 and 0.99.

On modeling and predicting popularity dynamics via integrating generative model and rich features

Understanding the mechanisms governing how an online message acquires more popularity than another, modeling how it gains popularity dynamically, and determining the method for predicting its dynamics popularity are of tremendous interest to related decision support systems. However, one major limitation of existing generative dynamics models is that the learning parameters are difficult to interpret and it is unclear whether it can be generalized for other messages, as they are trained for different messages independently and the feature data-based connections between messages are ignored. To alleviate the defects, we first perform experiments on real-world data from Sina Weibo to identify the general correlation between the dynamics model and rich features of online messages. Consequently, we present a novel feature-regularized dynamics model based on reinforced Poisson process (FRRPP), which regulates the parameter learning of popularity dynamics by integrating a feature regression term to capture the revealed correlation across online posts. Specifically, in addition to the objective of the maximum likelihood function, we assume that the competitiveness parameter of the different posts can be predicted by rich features, to enhance the explicability and generality of the point-process model and learn the dynamics process of different posts together. The proposed model is then evaluated on two real Sina Weibo datasets, and conclusive experimental results indicate that the proposed model achieves a remarkable improvement over baseline methods in terms of MAPE and Accuracy with various settings, which further verifies our findings about how to improve the generality of popularity dynamics modeling and prediction.

LSTM variants meet graph neural networks for road speed prediction

Traffic flow prediction is a fundamental issue in smart cities and plays an important role in urban traffic planning and management. An accurate predictive model can help individuals make reliable travel plans and choose optimal routes while efficiently helping administrators maintain traffic order. Road speed prediction, which is a sub-task of traffic flow forecasting, is challenging due to the complicated spatial dependencies characterizing road networks and dynamic temporal traffic patterns. Given the power of recurrent neural networks (RNNs) in learning temporal relations and graph neural networks (GNNs) in integrating graph-structured and node-attributed features, in this paper, we design a novel graph LSTM (GLSTM) framework to capture spatial-temporal representations in road speed forecasting. More specifically, we first present a temporal directed attributed graph to model complex traffic flow. Then, to take advantage of the structure properties and graph features, we employ a message-passing mechanism for feature aggregation and updating. Finally, we further implement several variants of LSTMs with a GN block under the encoder-decoder framework to model spatial-temporal dependencies. The experiments show that our proposed model is able to fully utilize both the road latent graph structure and traffic speed to forecast the road state during future periods. The results on two real-world datasets show that our GLSTM can outperform state-of-the-art baseline methods by up to 32.8% in terms of MAE, 43.2% in terms of MAPE and 23.1% in terms of RMSE.

A hybrid model of generalized regression neural network and radial basis function neural network for wind power forecasting in Indian wind farms

Wind power generation is a major part of renewable energy but it suffers from uncertainty and variable nature of wind speed. In conquering the uncertainty, forecasting of wind power generation emerges as best solution. Accordingly, wind power forecasting assists the power grid to increase the grid reliability. In this paper, a generalized regression neural network (GRNN), a neural network of radial basis function (RBFN) and a hybrid of GRNN and RBFN are applied for estimation of wind power and their performance is compared in respect to short-term wind power forecasting. In comparison to RBFN, GRNN has simple and static architecture. Historical wind power and meteorological wind speed data of the year 2014 from Indian wind farms are employed for training and testing the neural networks. The research work carried in this article interprets that GRNN has consistently performed better than RBFN. The hybrid GRNN-RBFN has also depicted good accuracy in a week ahead wind power forecasting. In such cases, either GRNN or RBFN predicts with high errors the hybrid GRNN-RBFN is helpful in accurate wind power forecasting. The errors are calculated in terms of MAPE & RMSE. For GRNN approach MAPE ranges from 0.48 % to 10.53 % and RMSE ranges from 0.01 KW to 3.31 KW. Whereas the maximum MAPE is 25.6 % for RBFN and the maximum RMSE is much higher i.e 11.61KW. In fact, for any forecasting approach, reliability is a major concern. The work carried also includes reliability analysis of all three forecasting models proposed. To ensure the best performance of GRNN, confidence intervals of MAPE are computed for all three models implemented. Among all, GRNN is well perceived to be the best one with narrowest confidence intervals.