Fuzzy-based Learning Research Articles

Ensemble Learning-Based Fuzzy Aggregation Functions and Their Application in TSK Neural Networks

AbstractTakagi–Sugeno–Kang fuzzy neural networks (TSKFNN) are powerful tools to model vague or imprecise information. Regression is one of the most important tasks commonly tackled by supervised learning techniques. TSKFNNs are considered suitable models to deal with regression problems on account of their simplicity and flexibility. Aggregation methods play an important role in combining various fuzzy rules from a TSKFNN rule base to obtain a model prediction. However, many current aggregation methods rely on expert experience and human knowledge, which may be hard to acquire and could bring human bias. This paper proposes data-driven aggregation functions for rules aggregation based on ensemble learning, namely AdaBoost and bagging, which can achieve superior generalizability in testing compared with the existing rule aggregation methods. Furthermore, they can also provide insights into the importance of each rule in the model’s decision making, thus, helping to improve the interpretability of the model. Extensive experiments on 11 commonly used benchmark datasets with various sizes and dimensionalities validated the superiority of the proposed ensemble learning-based fuzzy aggregation functions compared with existing state-of-the-art TSKFNNs.

Learning From Human Educational Wisdom: A Student-Centered Knowledge Distillation Method.

Existing studies on knowledge distillation typically focus on teacher-centered methods, in which the teacher network is trained according to its own standards before transferring the learned knowledge to a student one. However, due to differences in network structure between the teacher and the student, the knowledge learned by the former may not be desired by the latter. Inspired by human educational wisdom, this paper proposes a Student-Centered Distillation (SCD) method that enables the teacher network to adjust its knowledge transfer according to the student network's needs. We implemented SCD based on various human educational wisdom, e.g., the teacher network identified and learned the knowledge desired by the student network on the validation set, and then transferred it to the latter through the training set. To address the problems of current deficiency knowledge, hard sample learning and knowledge forgetting faced by a student network in the learning process, we introduce and improve Proportional-Integral-Derivative (PID) algorithms from automation fields to make them effective in identifying the current knowledge required by the student network. Furthermore, we propose a curriculum learning-based fuzzy strategy and apply it to the proposed PID control algorithm, such that the student network in SCD can actively pay attention to the learning of challenging samples after with certain knowledge. The overall performance of SCD is verified in multiple tasks by comparing it with state-of-the-art ones. Experimental results show that our student-centered distillation method outperforms existing teacher-centered ones.

Curriculum Learning-Based Fuzzy Support Vector Machine

Curriculum Learning-Based Fuzzy Support Vector Machine

Analysis of P2P Hosts Performance: An Integrated Deep Learning-Based Fuzzy Intuitionistic Approach

Analysis of P2P Hosts Performance: An Integrated Deep Learning-Based Fuzzy Intuitionistic Approach

A Cross-Field Deep Learning-based Fuzzy Spamming Detection Approach Via Collaboration of Behavior Modeling and Sentiment Analysis

A Cross-Field Deep Learning-based Fuzzy Spamming Detection Approach Via Collaboration of Behavior Modeling and Sentiment Analysis

An Adaptive Intrusion Detection System in the Internet of Medical Things Using Fuzzy-Based Learning.

The Internet of Medical Things (IoMT) is a growing trend within the rapidly expanding Internet of Things, enhancing healthcare operations and remote patient monitoring. However, these devices are vulnerable to cyber-attacks, posing risks to healthcare operations and patient safety. To detect and counteract attacks on the IoMT, methods such as intrusion detection systems, log monitoring, and threat intelligence are utilized. However, as attackers refine their methods, there is an increasing shift toward using machine learning and deep learning for more accurate and predictive attack detection. In this paper, we propose a fuzzy-based self-tuning Long Short-Term Memory (LSTM) intrusion detection system (IDS) for the IoMT. Our approach dynamically adjusts the number of epochs and utilizes early stopping to prevent overfitting and underfitting. We conducted extensive experiments to evaluate the performance of our proposed model, comparing it with existing IDS models for the IoMT. The results show that our model achieves high accuracy, low false positive rates, and high detection rates, indicating its effectiveness in identifying intrusions. We also discuss the challenges of using static epochs and batch sizes in deep learning models and highlight the importance of dynamic adjustment. The findings of this study contribute to the development of more efficient and accurate IDS models for IoMT scenarios.

On-Line learning-based Fuzzy Clustering Modelling in Real-Time Object Tracking

On-Line learning-based Fuzzy Clustering Modelling in Real-Time Object Tracking

Performance Improvement of Low-Cost Iterative Learning-Based Fuzzy Control Systems for Tower Crane Systems

This paper is dedicated to the memory of Prof. Ioan Dzitac, one of the fathers of this journal and its founding Editor-in-Chief till 2021. The paper addresses the performance improvement of three Single Input-Single Output (SISO) fuzzy control systems that control separately the positions of interest of tower crane systems, namely the cart position, the arm angular position and the payload position. Three separate low-cost SISO fuzzy controllers are employed in terms of first order discrete-time intelligent Proportional-Integral (PI) controllers with Takagi-Sugeno-Kang Proportional-Derivative (PD) fuzzy terms. Iterative Learning Control (ILC) system structures with PD learning functions are involved in the current iteration SISO ILC structures. Optimization problems are defined in order to tune the parameters of the learning functions. The objective functions are defined as the sums of squared control errors, and they are solved in the iteration domain using the recent metaheuristic Slime Mould Algorithm (SMA). The experimental results prove the performance improvement of the SISO control systems after ten iterations of SMA.

Reinforcement learning-based fuzzy geocast routing protocol for opportunistic networks

In a communication environment like opportunistic networks (OppNets) where there is no stable path, the message routing is a challenge. In this sense, using a deep learning approach that focuses on utilizing the agents in the environment to fulfill the message routing task has proven to be effective. This paper proposes a novel routing scheme for OppNets called Reinforcement Learning-based Fuzzy Geocast Routing Protocol (RLFGRP) for OppNets, in which a fuzzy controller makes use of the node’s Q-value, reward value and remaining buffer space as input parameters to determine the likelihood of that node to be selected as suitable forwarder of a message from source towards its destination. Through simulations using real mobility traces, the proposed RLFGRP protocol is shown to outperform the established Geocast Fuzzy-Based Check-and-Spray routing (FCSG) [4] and the Fuzzy logic-based Q-learning routing (FQLRP) [9] protocols in terms of overhead ratio, delivery ratio and average latency.

Detection of exon location in eukaryotic DNA using a fuzzy adaptive Gabor wavelet transform

The existing model-independent methods for the detection of exons in DNA could not prove to be ideal as commonly employed fixed window length strategy produces spectral leakage causing signal noise The Modified-Gabor-wavelet-transform exploits a multiscale strategy to deal with the issue to some extent. Yet, no rule regarding the occurrence of small and large exons has been specified. To overcome this randomness, scaling-factor of GWT has been adapted based on a fuzzy rule. Due to the nucleotides' genetic code and fuzzy behaviors in DNA configuration, this work could adopt the fuzzy approach. Two fuzzy membership functions (large and small) take care of the variation in the coding regions. The fuzzy-based learning parameter adaptively tunes the scale factor for fast and precise prediction of exons. The proposed approach has an immense plus point of being capable of isolating detailed sub-regions in each exon efficiently proving its efficacy comparing with existing techniques.

Global Exponential Stability of Learning-Based Fuzzy Networks on Time Scales

We investigate a class of fuzzy neural networks with Hebbian-type unsupervised learning on time scales. By using Lyapunov functional method, some new sufficient conditions are derived to ensure learning dynamics and exponential stability of fuzzy networks on time scales. Our results are general and can include continuous-time learning-based fuzzy networks and corresponding discrete-time analogues. Moreover, our results reveal some new learning behavior of fuzzy synapses on time scales which are seldom discussed in the literature.

Research on an Integrated Learning-based Fuzzy Cognitive Map Classifier Model

Research on an Integrated Learning-based Fuzzy Cognitive Map Classifier Model

A Fuzzy-Based Learning Vector Quantization Neural Network for Recurrent Nasal Papilloma Detection

The objective of this paper is to develop a complete solution for recurrent nasal papilloma (RNP) detection. Recently, Gadolinium-enhanced dynamic magnetic resonance imaging (MRI) has been developed and widely used in the clinical diagnosis of RNP. Because the response of RNP regions in Gadolinium-enhanced MR images is different from the response of normal tissues, the difference between the dynamic-MR images before and after administering a contrast material can be used to extract coarse RNP regions automatically. In this study, a fuzzy algorithm for learning vector quantization neural network is used to pick suspicious RNP regions. Finally, a feature-based region growing method is applied to recover complete RNP regions. The experimental results show that the proposed method can detect RNP regions automatically, correctly, and fast.

Fuzzy-based learning rate determination for blind source separation

Many independent component analysis (ICA) algorithms have been proposed for blind source separation. These algorithms belong to the LMS-type algorithm in natural. Hence, the choice of the step-size reflects a tradeoff between misadjustment and the speed of convergence. Based on the separation state of outputs of the neural network for ICA, the paper develops a fuzzy inference-based step-size selection algorithm. The fuzzy inference system consists of two inputs (the second- and higher order correlation coefficients of output components) and one output (the fuzzy learning rate). In this way, the ICA algorithms become more efficient, which is verified by simulation results.

A transformer resonant phase‐shifted PWM DC–DC converter for X‐ray power generator and its fuzzy logic‐based learning control scheme

AbstractIn general, medical‐use high‐voltage X‐ray power generators are operated at a variety of required output DC voltages and output current settings for the X‐ray tube. A phase‐shifted PWM inverter‐fed DC–DC power converter with a high‐voltage transformer parasitic resonant link which is used for an X‐ray power generator inherently has stiff nonlinear characteristics due to phase‐shifted voltage regulation and diode cutoff operation in a high‐voltage rectifier because of the wide load setting ranges in practical applications. However, the superior output voltage response required for a medical‐use X‐ray power generator could not be sufficiently implemented by various modern linear control approaches because a precise mathematical description with dynamic modeling could not be formulated for wide load variations. For controlling such a nonlinear DC–DC power conversion system, a fuzzy reasoning‐based learning control technique that duplicates human capabilities and much experience in understanding the dynamic and static behavior of the system are more suitable and effective. This paper presents a phase‐shifted PWM full bridge inverter‐type high‐voltage DC–DC power converter with a high‐voltage transformer parasitic link and its fuzzy‐based learning controller, which has excellent dynamic and static behavior, for use in a medical X‐ray power generator operating at a wide variety of load settings. The effectiveness of the feasible fuzzy learning technique used for control of the DC–DC power converter is discussed and evaluated based on computer simulations and experimental results in a laboratory setup. It is demonstrated that the feasible fuzzy learning control scheme suitable for this converter is more effective and practical in improving the system output voltage response performance in both transient and steady states for medical‐use X‐ray power generators. The results of simulations and experiments for this converter are presented in order to illustrate the operating performance of the proposed X‐ray power generator operated by the feasible fuzzy‐based learning control implementation. © 2001 Scripta Technica, Electr Eng Jpn, 136(2): 58–71, 2001

Transformer Resonant Phase-Shifted PWM DC-DC Converter for X Ray Power Generator and Its Fuzzy Logic-based Learning Control Scheme

In general, medical-use high-voltage X ray power generator works under a variety of operating conditions of the required output DC voltage and the output current setting values for the X ray tube. A phase-shifted PWM inverter-fed DC-DC power converter with a high-voltage transformer parasitic resonant link which is used for an X ray power generator inherently includes its stiff nonlinear characteristics due to phase-shifted voltage regulation and diode cut off operation in a high-voltage rectifier because of the wide load setting regulation ranges in practical applications. However, superior output voltage response performances of a quick rising time in a transient state and low ripple factor in a steady-state which have to be essentially required for a medical-use X ray power generator could not be sufficiently implemented by various linear modern control approaches because a precise mathematical description with dynamic modeling could not be formulated for wide load variations. For controlling such a nonlinear DC-DC power conversion system, a fuzzy-reasoning-based learning control technique to realize the human capability and much experience to understand the system dynamic and static behavior is more suitable and effective. This paper presents a phase-shifted PWM full bridge inverter type high-voltage DC-DC power converter with a high-voltage transformer parasitic resonant link and its fuzzy-based learning controller, which has excellent dynamic and static behaviors to be utilized in a medical-use X ray power generator operating under conditions of a variety of wide load settings. The effectiveness of feasible fuzzy learning control technique of the DC-DC power converter is discussed and evaluated on the basis of the computer simulation and experimental results in a laboratory set-up. It is proved that the feasible fuzzy-based learning control scheme suitable for this converter is more effective and practical to improve the system output voltage response performances in both the transient and steady states for medical-use X ray power generator. The simulated and experimental results of this converter are actually illustrated in order to substantiate the operating performance validity of the inverter type medical-use X ray power generator operated by the feasible fuzzy-based learning control implementation.