An End-to-End Adaptive Input Selection With Dynamic Weights for Forecasting Multivariate Time Series

Munkhdalai, Lkhagvadorj; Munkhdalai, Tsendsuren; Park, Kwang-Ho; Amarbayasgalan, Tsatsral; Erdenebaatar, Erdenebileg; Park, Hyun Woo; Ryu, Keun Ho

doi:10.1109/access.2019.2930069

Cited by 48 publications

(29 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…An End-to-End Adaptive Input Selection With Dynamic Weights for Forecasting Multivariate Time Series [18]: presents the Adaptive Input Selection with Recurrent Neural Network (AIS-RNN) for multivariate time series forecasting. The model consisted of two parts; the first model generated context-dependent importance weights for selecting proper inputs; afterwards, the second model based on the inputs predicted the target variable.…”

Section: Group 1: Neural Network (Nn)mentioning

confidence: 99%

Air Quality Prediction in Smart Cities Using Machine Learning Technologies Based on Sensor Data: A Review

2020

View full text Add to dashboard Cite

The influence of machine learning technologies is rapidly increasing and penetrating almost in every field, and air pollution prediction is not being excluded from those fields. This paper covers the revision of the studies related to air pollution prediction using machine learning algorithms based on sensor data in the context of smart cities. Using the most popular databases and executing the corresponding filtration, the most relevant papers were selected. After thorough reviewing those papers, the main features were extracted, which served as a base to link and compare them to each other. As a result, we can conclude that: (1) instead of using simple machine learning techniques, currently, the authors apply advanced and sophisticated techniques, (2) China was the leading country in terms of a case study, (3) Particulate matter with diameter equal to 2.5 micrometers was the main prediction target, (4) in 41% of the publications the authors carried out the prediction for the next day, (5) 66% of the studies used data had an hourly rate, (6) 49% of the papers used open data and since 2016 it had a tendency to increase, and (7) for efficient air quality prediction it is important to consider the external factors such as weather conditions, spatial characteristics, and temporal features.

show abstract

Section: Group 1: Neural Network (Nn)mentioning

confidence: 99%

Air Quality Prediction in Smart Cities Using Machine Learning Technologies Based on Sensor Data: A Review

2020

View full text Add to dashboard Cite

show abstract

“…The proposed hybrid CNN-GRU model was also compared with other baseline models by performing experiments on AEP [45] dataset. These models include XGBoost [51], Gradient boosting machine [52] and AIS-RNN [53]. The comparison of our proposed model using different evaluation metrics with…”

Section: E Comparison Of Proposed Hybrid Cnn-gru Model Over Aep Datamentioning

confidence: 99%

A Novel CNN-GRU-Based Hybrid Approach for Short-Term Residential Load Forecasting

et al. 2020

View full text Add to dashboard Cite

Electric energy forecasting domain attracts researchers due to its key role in saving energy resources, where mainstream existing models are based on Gradient Boosting Regression, Artificial Neural Networks, Extreme Learning Machine and Support Vector Machine. These models encounter high-level of non-linearity between input data and output predictions and limited adoptability in real-world scenarios. Meanwhile, energy forecasting domain demands more robustness, higher prediction accuracy and generalization ability for real-world implementation. In this paper, we achieve the mentioned tasks by developing a hybrid sequential learning-based energy forecasting model that employs Convolution Neural Network and Gated Recurrent Units into a unified framework for accurate energy consumption prediction. The proposed framework has two major phases: (1) data refinement and (2) training, where the data refinement phase applies preprocessing strategies over raw data. In the training phase, CNN features are extracted from input dataset and fed in to GRU, that is selected as optimal and observed to have enhanced sequence learning abilities after extensive experiments. The proposed model is an effective alternative to the previous hybrid models in terms of computational complexity as well prediction accuracy, due to the representative features' extraction potentials of CNNs and effectual gated structure of multi-layered GRU. The experimental evaluation over existing energy forecasting datasets reveal the better performance of our method in terms of preciseness and efficiency. The proposed method achieved the smallest error rate on individual Appliances Energy prediction and household electric power consumption datasets, when compared to other baseline models.

show abstract

“…Chang et al [43] proposed the MTNet model, which includes a large memory component composed of multiple LSTMs, three independent coders and attention mechanism to capture spatial and temporal features. Munkhdalai et al [44] proposed AIS-RNN framework, which combines RNNs with an adaptive input selection mechanism to improve prediction performance.…”

Section: A Related Workmentioning

confidence: 99%

A Hybrid Method With Adaptive Sub-Series Clustering and Attention-Based Stacked Residual LSTMs for Multivariate Time Series Forecasting

Liu

Cai

2020

IEEE Access

View full text Add to dashboard Cite

Multivariate Time Series Forecasting (MTSF) has recently emerged its growing importance in many industries. However, how to reduce the influence of the noise components existing in time series on prediction and extract features effectively are still two challenges in MTSF. This paper focuses on those two challenges and proposes a new prediction method based on decomposition-ensemble framework called adaptive sub-series clustering-stacked residual LSTMs-multi-level attention mechanism (ASC-SRLSTMs-MLAttn). The method consists of three stages: decomposition, prediction, and ensemble. In the decomposition stage, the target series is decomposed by Ensemble Empirical Mode Decomposition into multiple sub-series, which will be clustered and reconstructed by the ASC algorithm to reduce the complexity and the time consumption of prediction. In the prediction stage, the sub-series and correlation series will be fed into SRLSTMs-MLAttn for sub-series prediction. The model is based on the encoder-decoder architecture with stacked residual LSTMs as the encoder, which can effectively capture the dependencies among multi variables and the temporal features from multivariate time series. Besides, a multi-level attention mechanism (MLAttn), which makes full use of the encoding information of the encoder, has been introduced to further improve the prediction performance of the model. In the ensemble stage, the predicted values of each sub-series will be summed to obtain the final prediction of the target time series. We also demonstrate the superiority and effectiveness of our proposed method on four public datasets via the conducted comparison experiment and ablation study. INDEX TERMS Multivariate time series forecasting, decomposition-ensemble framework, adaptive sub-series clustering algorithm, stacked residual long short-term memory, multi-level attention mechanism.

show abstract

An End-to-End Adaptive Input Selection With Dynamic Weights for Forecasting Multivariate Time Series

Cited by 48 publications

References 45 publications

Air Quality Prediction in Smart Cities Using Machine Learning Technologies Based on Sensor Data: A Review

Air Quality Prediction in Smart Cities Using Machine Learning Technologies Based on Sensor Data: A Review

A Novel CNN-GRU-Based Hybrid Approach for Short-Term Residential Load Forecasting

A Hybrid Method With Adaptive Sub-Series Clustering and Attention-Based Stacked Residual LSTMs for Multivariate Time Series Forecasting

Contact Info

Product

Resources

About