Scopus İndeksli Yayınlar Koleksiyonu / Scopus Indexed Publications Collection

Permanent URI for this collectionhttps://hdl.handle.net/11147/7148

Browse

Filters

2018 - 201912020 - 20211

Settings

Publisher: search.filters.publisher.IEEESubject: search.filters.subject.Neural networks

Search Results

Now showing 1 - 2 of 2
  • Article
    Citation - WoS: 8
    Citation - Scopus: 14
    An End-To Trainable Feature Selection-Forecasting Architecture Targeted at the Internet of Things
    (IEEE, 2021) Nakıp, Mert; Karakayalı, Kubilay; Güzeliş, Cüneyt; Rodoplu, Volkan
    We develop a novel end-to-end trainable feature selection-forecasting (FSF) architecture for predictive networks targeted at the Internet of Things (IoT). In contrast with the existing filter-based, wrapper-based and embedded feature selection methods, our architecture enables the automatic selection of features dynamically based on feature importance score calculation and gamma-gated feature selection units that are trained jointly and end-to-end with the forecaster. We compare the performance of our FSF architecture on the problem of forecasting IoT device traffic against the following existing (feature selection, forecasting) technique pairs: Autocorrelation Function (ACF), Analysis of Variance (ANOVA), Recurrent Feature Elimination (RFE) and Ridge Regression methods for feature selection, and Linear Regression, Multi-Layer Perceptron (MLP), Long Short Term Memory (LSTM), 1 Dimensional Convolutional Neural Network (1D CNN), Autoregressive Integrated Moving Average (ARIMA), and Logistic Regression for forecasting. We show that our FSF architecture achieves either the best or close to the best performance among all of the competing techniques by virtue of its dynamic, automatic feature selection capability. In addition, we demonstrate that both the training time and the execution time of FSF are reasonable for IoT applications. This work represents a milestone for the development of predictive networks for IoT in smart cities of the near future.
  • Article
    Citation - WoS: 9
    Citation - Scopus: 11
    Neural Network-Based Repetitive Learning Control of Euler Lagrange Systems: an Output Feedback Approach
    (IEEE, 2018) Tatlıcıoğlu, Enver; Çobanoğlu, Necati; Zergeroǧlu, Erkan
    In this letter, position tracking control problem of a class of fully actuated Euler Lagrange (EL) systems is aimed. The reference position vector is considered to be periodic with a known period. Only position measurements are available for control design while velocity measurements are not. Furthermore, the dynamic model of the EL systems has parametric and/or unstructured uncertainties which avoid it to be used as part of the control design. To address these constraints, an output feedback neural network-based repetitive learning control strategy is preferred. Via the design of a dynamic model independent velocity observer, the lack of velocity measurements is addressed. To compensate for the lack of dynamic model knowledge, universal approximation property of neural networks is utilized where an online adaptive update rule is designed for the weight matrix. The functional reconstruction error is dealt with the design of a novel repetitive learning feedforward term. The outcome is a dynamic model independent output feedback neural network-based controller with a repetitive learning feedforward component. The stability of the closed-loop system is investigated via rigorous mathematical tools with which semi-global asymptotic stability is ensured. © 2017 IEEE.