Neural network architecture based on gradient boosting for IoT traffic prediction

López-Martín, Manuel; Carro, Belén; Sánchez-Esguevillas, Antonio

doi:10.1016/j.future.2019.05.060

Cited by 45 publications

(65 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The naïve Bayes algorithm fails to classify the incoming traffic when the network traffic is high. Martin et al [19] proposed a deep learning-based neural network architecture for IoT traffic prediction. This architecture contains residual, boosted, and stacked networks.…”

Section: Traffic Differentiation Schemesmentioning

confidence: 99%

AI-aided Traffic Differentiated QoS Routing and Dynamic Offloading in Distributed Fragmentation Optimized SDN-IoT

Begovic¹,

Čaušević²,

Memić³

et al. 2020

International Journal of Engineering Research and Technology

View full text Add to dashboard Cite

Internet of Things (IoT) becomes an emerging network technology that expedites billions of devices to be connected via the Internet to provide real-time intelligent application services. The benefits of Software-Defined Networking (SDN) can be used to fulfill IoT requirements. Quality of Service provisioning is an ongoing demand in software-defined IoT (SD-IoT), particularly for large scale environments. In this paper, we address this issue by proposing a seamless model of AI-aided Traffic Differentiated QoS Routing and Dynamic Offloading in distributed fragmentation optimized SDN-IoT. Firstly, we propose a Multi-Criterion based Deep Packet Inspection method for classifying the network traffic, which is held in Edge Routers (access points). Secondly, we construct a Partially Connected Network Topology using the ISOMAP algorithm for an effective rule placement and routing. We propose a Traffic Differentiated QoS Routing for forwarding data packets via the most suitable switches. We select the optimum route by Deep Alternative Neural Network (DANN). Based on the relationships among switches, the path is selected and flow rules are deployed. The poor QoS is often caused by load imbalance in controllers and switches. To overwhelm this issue, we propose a Dynamic Offloading scheme in SD-IoT. We offload the data packets from the overloaded controller to the underloaded controller using Hassanat Distance-based Knearest neighbors (HDK-NN) algorithm. Similarly, we propose a Ranking-based Entropy function (R-Ef) to allow dynamic offloading among switches. Simulation is performed using the NS3.26 simulator and the results proved that our proposed AIaided SD-IoT model provides superior QoS performance compared to previous approaches.

show abstract

Section: Traffic Differentiation Schemesmentioning

confidence: 99%

AI-aided Traffic Differentiated QoS Routing and Dynamic Offloading in Distributed Fragmentation Optimized SDN-IoT

Begovic¹,

Čaušević²,

Memić³

et al. 2020

International Journal of Engineering Research and Technology

View full text Add to dashboard Cite

show abstract

“…Finally, the information to consider from the previous time-slots (predictors) can be extended to the load values and additional information, such as date/time or weather data (multivariate forecast). Including this additional (exogenous) information as new features imposes difficulties on statistical analysis methods since only a few can cope with multivariate forecasts [ 9 ]. This creates additional opportunities for machine learning and deep learning (ML/DL) techniques that easily handle vector-valued predictors.…”

Section: Introductionmentioning

confidence: 99%

“…The only requirement for a base model is to be trainable end-to-end by gradient descent and support the addition of a final layer in both the training and prediction stages. Thus, we have considered as base models several configurations of 1D and 2D convolutional neural networks (CNN) [ 13 , 14 ], long short-term memory (LSTM) [ 15 ] networks and their combination, as well as several additive ensembles (AE) deep learning models especially suitable for time-series forecasting [ 9 , 16 ]. We do not include sequence-to-sequence (Seq2seq) models as a base model since the forward pass for the training, and test stages are different with added complexity for the proposed extension.…”

Section: Introductionmentioning

confidence: 99%

“…We provide a comprehensive comparison between CWQFNN and a significant number of state-of-the-art (SOTA) data-driven forecasting models, some of them widely applied to time-series forecasting and others novel or rarely applied to SMTLF, such as (a) classic machine learning (ML) models, e.g., linear regression and random forest [ 18 , 19 , 20 , 21 ], (b) multilayer perceptron [ 5 ], (c) deep learning models based on separate CNN and recurrent neural networks (RNN) [ 19 , 20 ], (d) dynamic mode decomposition (DMD) [ 22 , 23 , 24 ], (e) deep learning (DL) models based on specific combinations of CNN and RNN [ 25 , 26 ], (f) sequence-to-sequence (Seq2seq) models with and without soft attention [ 27 , 28 , 29 ], and (g) deep learning additive ensemble models especially targeted for time-series forecasting [ 9 ].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Additive Ensemble Neural Network with Constrained Weighted Quantile Loss for Probabilistic Electric-Load Forecasting

López-Martín

Sánchez-Esguevillas

Hernández-Callejo

et al. 2021

Sensors

Self Cite

View full text Add to dashboard Cite

This work proposes a quantile regression neural network based on a novel constrained weighted quantile loss (CWQLoss) and its application to probabilistic short and medium-term electric-load forecasting of special interest for smart grids operations. The method allows any point forecast neural network based on a multivariate multi-output regression model to be expanded to become a quantile regression model. CWQLoss extends the pinball loss to more than one quantile by creating a weighted average for all predictions in the forecast window and across all quantiles. The pinball loss for each quantile is evaluated separately. The proposed method imposes additional constraints on the quantile values and their associated weights. It is shown that these restrictions are important to have a stable and efficient model. Quantile weights are learned end-to-end by gradient descent along with the network weights. The proposed model achieves two objectives: (a) produce probabilistic (quantile and interval) forecasts with an associated probability for the predicted target values. (b) generate point forecasts by adopting the forecast for the median (0.5 quantiles). We provide specific metrics for point and probabilistic forecasts to evaluate the results considering both objectives. A comprehensive comparison is performed between a selection of classic and advanced forecasting models with the proposed quantile forecasting model. We consider different scenarios for the duration of the forecast window (1 h, 1-day, 1-week, and 1-month), with the proposed model achieving the best results in almost all scenarios. Additionally, we show that the proposed method obtains the best results when an additive ensemble neural network is used as the base model. The experimental results are drawn from real loads of a medium-sized city in Spain.

show abstract

“…It has been shown through various studies [8]- [11] that the accuracy and stability of the model can be improved by combining the results of the predictive models. Boosting, proposed by Freund [12], improves performance by combining multiple weak models (i.e., submodels) into a strong model unlike bagging [13], [14]. The principal idea of boosting is to generate models sequentially and to improve overall model performance thereby.…”

Section: Introductionmentioning

confidence: 99%

Additive Ensemble Neural Networks

et al. 2020

View full text Add to dashboard Cite

Deep neural networks (DNNs) have been making progress in many ways. DNNs are typically used to model complex nonlinearity of high-dimensional data in regression or classification problems. As DNNs contain additional hidden layers, they generally improve performance but increase the number of parameters to train, thereby extending the learning time. Many studies, such as those employing Dropout and regularization methods, have undertaken to solve these problems. The method proposed in this paper is an additive ensemble neural networks (AENNs), by which a boosting mechanism of an ensemble methodology is applied to the neural networks instead of regularization techniques. That is, the model by AENNs is obtained by sequentially combining several simple shallow network models. Experiments showed that AENNs yield better results than conventional DNNs and machine learning methods for regression and classification problems, thereby alleviating the troublesome model complexity issue.

show abstract

Neural network architecture based on gradient boosting for IoT traffic prediction

Cited by 45 publications

References 46 publications

AI-aided Traffic Differentiated QoS Routing and Dynamic Offloading in Distributed Fragmentation Optimized SDN-IoT

AI-aided Traffic Differentiated QoS Routing and Dynamic Offloading in Distributed Fragmentation Optimized SDN-IoT

Additive Ensemble Neural Network with Constrained Weighted Quantile Loss for Probabilistic Electric-Load Forecasting

Additive Ensemble Neural Networks

Contact Info

Product

Resources

About