If-then Rules Research Articles

Clustering method for production of Z-number based if-then rules

Application of clustering algorithms to extract or summarize data from large data sets is a straightforward and effective approach. Sometimes data sets are not only large in size but also include imprecise and partially reliable pieces of information of probabilistic and possibilistic (fuzzy) nature. Information of this kind is referred to as bimodal information. There is a need to summarize such data into a compact set of human tractable “If-Then” rules, being a result of synergy of imprecision of both types. In this paper, we first present an approach to clustering with the purpose of extraction of probabilistic and fuzzy information (bimodal) categories suitable for formation of human-tractable rules. The suggested clustering algorithm utilizes the FCM objective function and an evolutionary algorithm to produce fuzzy sets of fuzzy clusters. To form Z-clusters and use them as components for the rules we exploit the relationship existing between Type-2 Fuzzy and Z-number concepts. A benchmark problem and a real-world application are considered to demonstrate the usefulness of the proposed approach.

Robotic architecture as unknown discrete-time system based on variable-frequency drive and adaptive controller

Abstract The new architecture for outdated robotic arm (ABB: IRB-1400) is developed by the direct variable-frequency drive (VFD) and considered as a class of unknown nonlinear discrete-time systems. The identification model is established by Multi-input fuzzy rules emulated network (MiFREN) with the online learning algorithm and the database of IF-THEN rules. The controller is designed by using the identification model and feedback signals from the multiplexer when kinematic and dynamic models of the robotic system are neglected. The convergence of tracking errors and internal signals is guaranteed by both theoretical and practical aspects. The performance validation of the proposed robotic system is accomplished by the experimental setup with comparative results.

An interpretable fuzzy logic based data-driven model for the twin screw granulation process

In this research, a new framework based on fuzzy logic is proposed to model the twin screw granulation (TSG) process. First, various fuzzy logic systems (FLSs) having different structures are developed to define various rule bases. The extracted fuzzy rules are assessed and reduced accordingly into a single rule base by utilizing the singular value decomposition-QR factorization (SVD-QR) approach. The resulted reduced FLS is, then, implemented to describe the TSG process mathematically and linguistically via simple to understand IF-THEN rules. The linguistic output provides an accessible framework to increase the understanding of this complex process within an industrial context. Validated on laboratory-scale experiments, it is shown that the newly proposed model successfully predicts the granule size and enhances the understanding of the TSG process. Furthermore, the proposed framework outperforms the standard FLS and the Artificial Neural Network (ANN), with an overall improvement of approximately 16% and 29% in R2, respectively.

Optimal Longitudinal Control for Vehicular Platoon Systems: Adaptiveness, Determinacy, and Fuzzy

This article addresses the control design problem for a fuzzy vehicular platoon system consisting of one leading vehicle and N following vehicles. Due to external disturbances, uncertainty exists in the platoon system and is usually nonlinear and possibly fast time varying. In order to deal with uncertainty, fuzzy theory is employed to describe the platoon system, thereby the so-called fuzzy vehicular platoon system. Based on the fuzzy property of uncertainty, a type of adaptive law is proposed. Since the original state of the system is one-side bounded when collision avoidance is taken into account, a state transformation is made, by which a new global state is obtained. Then, the swarm intelligence is embedded into the platoon system via a function, which mimics the swarm behavior. For the control design, we focus on adaptiveness and optimization. A switching-type adaptive robust control is proposed. The control is not IF-THEN rule based. We further explore the problem of control parameter optimization. A performance index consisting of transient control cost and average control cost is proposed. Aiming at minimizing the performance index, the optimal problem is formulated. The solution to the optimal problem (i.e., optimal control parameter) exists and is unique. The control with optimal parameter is called optimal control, which guarantees both deterministic performance (uniform boundedness, uniform ultimate boundedness, and collision avoidance) and fuzzy performance of the platoon system.

Robust control design of electric helicopter tail reduction system: Fuzzy and optimal view

The optimal robust control with a fuzzy approach is applied to design the electric helicopter tail reduction system in this article. Firstly, the fuzzy dynamical model of the electric helicopter tail reduction system with uncertainties and external disturbances is established, which may be time varying. Then, we propose a deterministic robust controller (differing from IF-THEN rules in traditional fuzzy control) to solve the uncertain problem in electric helicopter tail reduction systems. The electromechanical system with the proposed controller is proved to be stable via the Lyapunov function. The control gain with an optimal design is considered, which minimizes a fuzzy performance index associated with both the control error and the control cost. Furthermore, simulations compared with linear-matrix-inequality control are made to validate the effectiveness and stability of the optimal robust controller. All results show that the performance of the fuzzy electric helicopter tail reduction system can be always guaranteed by this optimal robust control approach.

Entropy clustering-based granular classifiers for network intrusion detection

Support vector machine (SVM) is one of the effective classifiers in the field of network intrusion detection; however, some important information related to classification might be lost in the reprocessing. In this paper, we propose a granular classifier based on entropy clustering method and support vector machine to overcome this limitation. The overall design of classifier is realized with the aid of if-then rules that consists of a premise part and conclusion part. The premise part realized by the entropy clustering method is used here to address the problem of a possible curse of dimensionality, while the conclusion part realized by support vector machines is utilized to build local models. In contrast to the conventional SVM, the proposed entropy clustering-based granular classifiers (ECGC) can be regarded as an entropy-based support function machine. Moreover, an opposition-based genetic algorithm is proposed to optimize the design parameters of the granular classifiers. Experimental results show the effectiveness of the ECGC when compared with some classical models reported in the literatures.

A multicriteria forecast of the default probability in credit risk assessment

This study analyses, through a multicriteria approach producing if-then rules, the possibility of forecasting the default probability (PD) for low-default portfolios in credit risk assessment. This approach relies on the simultaneous and complementary application of fuzzy and rough sets theories. The calculation of this probability is normally subject to judgments of experts, which imply subjectivity, imprecision and uncertainty of the results. In the proposed approach, the inferred if-then rules for a decision table generated by a simulation of 50 companies allowed the production of a knowledge base for the forecasting of classes of default probabilities. The 'RoughSets®' package, in 'R', was used as support for the analysis and forecasting system.

Detection of waterborne bacteria using Adaptive Neuro-Fuzzy Inference System

The detection of waterborne bacteria is crucial to prevent health risks. Current research uses soft computing techniques based on Artificial Neural Networks (ANN) for the detection of bacterial pollution in water. The limitation of only relying on sensor-based water quality analysis for detection can be prone to human errors. Hence, there is a need to automate the process of real-time bacterial monitoring for minimizing the error, as mentioned above. To address this issue, we implement an automated process of water-borne bacterial detection using a hybrid technique called Adaptive Neuro-fuzzy Inference System (ANFIS), that integrates the advantage of learning in an ANN and a set of fuzzy if-then rules with appropriate membership functions. The experimental data as the input to the ANFIS model is obtained from the open-sourced dataset of government of India data platform, having 1992 experimental laboratory results from the years 2003-2014. We have included the following water quality parameters: Temperature, Dissolved Oxygen (DO), pH, Electrical conductivity, Biochemical oxygen demand (BOD) as the significant factors in the detection and existence of bacteria. The membership function changes automatically with every iteration during training of the system. The goal of the study is to compare the results obtained from the three membership functions of ANFIS- Triangle, Trapezoidal, and Bell-shaped with 35 = 243 fuzzy set rules. The results show that ANFIS with generalized bell-shaped membership function is best with its average error 0.00619 at epoch 100.

Neuro-Fuzzy Networks and Their Applications in Medical Fields.

The fusion of artificial neural networks and fuzzy logic systems allows researchers to model real-world problems through the development of intelligent and adaptive systems. Artificial neural networks are able to adapt and learn by adjusting the interconnections between layers, while fuzzy logic inference systems provide a computing framework based on the concept of fuzzy set theory, fuzzy if-then rules, and fuzzy reasoning. The combined use of those adaptive structures is known as "neuro-fuzzy" systems. In this paper, the basic elements of both approaches are analyzed, noticing that this blending could be applied for pattern recognition in medical applications.

Disturbance Observer-Based Linear Matrix Inequality for the Synchronization of Takagi-Sugeno Fuzzy Chaotic Systems

This article presents a synchronization control method based on poles’ placement, disturbances, and uncertainty estimation (DUE) for a pair of Takagi-Sugeno fuzzy systems. First, a 3-D chaotic system was completely converted into a Takagi-Sugeno (T-S) fuzzy model by applying the nonlinearity sector method, which consists of if-then rules and sub-linear systems. Second, two identical T-S fuzzy systems with different initial conditions were synchronized by applying the linear matrix inequality (LMI) to place the eigenvalues of the state error equations in the stable region. Third, the sum of the time-varying disturbances and uncertainties of two nonidentical T-S fuzzy systems were deleted by a disturbance and uncertainty estimation. The given output signals confirmed that the proposed method is suitable and ideal for synchronizing T-S fuzzy systems. The ideas of control theory were implemented by using two experimental scenarios in MATLAB Simulink for two computers connected via an internet router and an electronics circuit’s communication.

Hot News Detection System Using Rule-Based Technique

Big data analytics provides the large volumes of data. The data are collected from a various sources including social networks, videos, digital images, sensors, and sales transaction records. Hot news is the information about current news of anywhere in the world. The information is receiving about an event that just happened. Breaking news reported while an event is on-going. In existing system,used CNN based hash tag recommendation along with the increasing requirements. Convolutionalneural networks (CNNs) for many natural language processing tasks, to perform the hashtag recommendation problem. In this work, proposedan efficient rule-based technique for merging news and social streams in real-time, Rule-based classifier makes use of a set of IF-THEN rules for classification. Twitter hash tags function as a key connection between Twitter crowds and the news media. To improve the efficiency and coverage of a state-of-the-art hash tag recommendation model by proposing new techniques for data collection and feature computation

Developing Decision Tree based Models in Combination with Filter Feature Selection Methods for Direct Marketing

Direct Marketing is a form of advertising strategies which aims to communicate directly with the most potential customers for a certain product using the most appropriate communication channel. Banks are spending a huge amount of money on their marketing campaigns, so they are increasingly interested in this topic in order to maximize the efficiency of their campaigns, especially with the existence of high competition in the market. All marketing campaigns are highly dependent on the huge amount of available data about customers. Thus special Data Mining techniques are needed in order to analyze these data, predict campaigns efficiency and give decision makers indications regarding the main marketing features affecting the marketing success. This paper focuses on four popular and common Decision Tree (DT) algorithms: SimpleCart, C4.5, RepTree and Random Tree. DT is chosen because the generated models are in the form of IF-THEN rules which are easy to understand by decision makers with poor technical background in banks and other financial institutions. Data was taken from a Portuguese bank direct marketing campaign. A filter-based Feature selection is applied in the study to improve the performance of the classification. Results show that SimpleCart has the best results in predicting the campaigns success. Another interesting finding that the five most significant features influencing the direct marketing campaign success to be focused on by decision makers are: Call duration, offered interest rate, number of employees making the contacts, customer confidence and changes in the prices levels.

Fuzzy logic model for stability assessment of underground facilities

A proper assessment of the stability leads to proactive decision making for the method of construction, the shape and size of the cross-section of the facility and the adequate support. The stability is directly dependent on the physical, mechanical and technical characteristics of the rock mass. The traditional assessment-based approach and the classification of stability of underground facilities is based on both experience and experiments with previously constructed facilities. To ease the process for the early phase of design, there was a need for the development of a mathematical model for assessment of the stability. Ambiguity in mining engineering, directs the development of the model using a fuzzy logic system and fuzzy if-then rules. The model was developed based on very scarce input data that is most often the only data available during the design stage. Fuzzy model results have shown, through the random selection of input data in the model, that in 70% of cases, there is high numerical congruence with the traditional method for determination of coefficient of stability S; and in 90% of cases, there is also congruence in accordance with the traditional approach to the classification of underground facilities based on the stability assessment.

New unified insights on deep learning in radiological and pathological images: Beyond quantitative performances to qualitative interpretation

Abstract Deep learning (DL) has become the main focus of research in the field of artificial intelligence, despite its lack of explainability and interpretability. DL mainly involves automated feature extraction using deep neural networks (DNNs) that can classify radiological and pathological images. Convolutional neural network (CNNs) can be also applied to pathological image analysis, such as the detection of tumors and the quantification of cellular features. However, to our knowledge, no attempts have been made to identify interpretable signatures from CNN features, and few studies have examined the use of CNNs for cytopathology images. Therefore, the aim of the present paper is to provide new unified insights to aid the development of more interpretable CNN-based methods to classify radiological and pathological images and explain the reason for this classification in the form of if-then rules. We first describe the “black box” problem of shallow NNs, the concept of rule extraction, the renewed attack of the “black box” problem in DNN architectures, and the paradigm shift regarding the transparency of DL using rule extraction. Next, we review limitations of DL in pathology in regard to histopathology and cytopathology. We then investigate the discrimination of cytological features and explanations and review recent techniques for interpretable CNN-based methods in histopathology, as well as current approaches being taken to enhance the interpretability of CNN-based methods for radiological images. Finally, we provide new unified insights to extract qualitative interpretable rules for radiological and pathological images.

Knowledge-based preference learning model for recommender system using adaptive neuro-fuzzy inference system

A recommender system (RS) delivers personalized suggestions on products based on the interest of a particular user. Content-based filtering (CBF) and collaborative filtering (CF) schemes have been previously used for this task. However, the main challenge in RS is cold start problem (CSP). This originates once a new user joins the system which makes the recommendation task tedious due to the shortage of information (clickstream, dwell time, rating, etc.) regarding the user’s interest. Therefore, CBF and CF are combined together by developing a knowledge-based preference learning (KBPL) system. This system considers the demographic data that includes gender, occupation, and age for the recommendation task. Initially, the dataset is clustered using the self-organizing map (SOM) technique, then the high dimensional data is decomposed by higher-order singular value decomposition (HOSVD) and finally, Adaptive neuro-fuzzy inference system (ANFIS) predicts the output. For the big dataset, SOM is a robust clustering method and the similarities among the users can be easily observed by grid clustering. The HOSVD extracts the required information from the available data set to find the user similarity by decomposing the dataset in lower dimensions. ANFIS uses IF-THEN rules to recommend similar product to the new users. The proposed KBPL system is evaluated with the Black Friday dataset and the obtained error value is compared with the existing CF and CBF techniques. The proposed KBPL system has obtained root mean squared error (RMSE) of 0.71%, mean absolute error (MAE) of 0.54%, and mean absolute percentage error (MAPE) of 37%. Overall, the outcome of the comparative analysis shows minimum error and better performance in terms of precision, recall, and f-measure for the proposed KBPL system compared to the existing techniques and therefore more suitable for accurately recommending the products for the new users.

Bi-level optimization based on fuzzy if-then rule

A bi-level programming problem has been developed where the functional relationship linking decision variables and the objective functions of leader and follower are not utterly well known to us. Because of the uncertainty in practical life decision-making situation most of the time it is inconvenient to find the veracious relationship between the objective functions of leader, follower and the decision variables. It is expected that the source of information which gives some command about the objective functions of leader and follower, is composed by a block of fuzzy if-then rules. In order to analyze the model, A dynamic programming approach with a suitable fuzzy reasoning scheme is applied to calculate the deterministic functional relationship linking the decision variables and the objective functions of leader as well as follower. Thus a bi-level programming problem is constructed from the actual fuzzy rule-based to the conventional bi-level programming problem. To solve the final problem, we use the lingo software to find the optimal of objective function of follower first and using its solution we optimize the objective function of leader. A numerical example has been solved to signify the computational procedure.

Novel similarity measure-based random forest for fingerprint recognition using dual-tree complex wavelet transform and ring projection

Designing an efficient fingerprint recognition technique is an ill-posed problem. Recently, many researchers have utilized machine learning techniques to improve the fingerprint recognition rate. The random forest (RF) is found to be one of the extensively utilized machine learning techniques for fingerprint recognition. Although it provides good recognition results at significant computational speed, still there is room for improvement. RF is not so-effective for high-dimensional features and also when features contain both discrete and continuous values at the same time. Therefore, in this paper, a novel similarity measure-based random forest (NRF) is proposed. The proposed technique, initially, computes both mutual information and conditional entropy. Thereafter, it uses three designed if-then rules to obtain final information measure. Additionally, to obtain feature set for fingerprint dataset, dual-tree complex wavelet transform is used to evaluate complex detail coefficients. Thereafter, ring project is considered to compute significant moments from these complex detail coefficients. Finally, information gain-based feature selection technique is used to select potential features. To prevent over-fitting, 20-fold cross validation is also used. Extensive experiments are considered to evaluate the effectiveness of the proposed technique. The comparative analyses reveal that the proposed technique outperforms the existing techniques in terms of accuracy, f-measure, sensitivity, specificity, kappa statistics and computational speed.

A Classification of Internet Pornographic Images

According to Pornography Statistics,more than 34 percent of Internet users exposeto pornography. There are 12 percent of the total number of websites and 72 million monthly visitors.Internet pornography (Internet Porn) is addictive to teenagers and kids around the world. The normal practice is to block those websites or filter out pornographyfrom kids.In order to do so, researchers has to find a way to detect and classify first. The pixel features including YCbCr range, area of human skin are chosen as pornographyfeatures because of their easy acquisition. C4.5 (Data mining technique)is applied to construct a decision tree. The purpose of this paper is to classify pornography images in a simple if-then rule. The accuracy of experimental result is 85.2%.

전반부 불일치 조건에서의 불확실한 시스템의 상태 추정을 위한 샘플데이터 퍼지 관측기 설계

In this paper, we propose a sampled-data fuzzy observer design technique for estimating the state variables of a nonlinear system with model uncertainty. It is assumed that the IF-THEN rules of the fuzzy system contains immeasurable premise variables, which complicates the observer design. In this paper, the observer is assumed not to share the same premise part with that of the system in order to deal with the immeasurable premise condition. After then, the error between the observer and the system including model uncertainty is represented by the Takagi-Sugeno (T-S) fuzzy model. In order to minimize the impact of imperfect premise matching, model uncertainty, and disturbances on the state estimation, an HSUB∞/SUB performance criterion is defined. Based on the fuzzy Lyapunov function, we derive a sufficient condition in the form of linear matrix inequality to ensure that the error dynamics is asymptotically stable and satisfy the HSUB∞/SUB condition. Finally, a simulation example verifies the superiority of the proposed method.

Improvement of the ANFIS-based wave predictor models by the Particle Swarm Optimization

In this paper, the Particle Swarm Optimization (PSO) algorithm is employed to deal with the Adaptive Network based Fuzzy Inference System (ANFIS) model drawbacks in prediction of wind –driven waves. In the ANFIS model selection of fuzzy IF-THEN rules structure and numbers is not an automatic process. In addition, in the ANFIS model extraction of fuzzy antecedent and consequent parameters is a gradient-based method which makes the answer susceptible to entrap in local optima. To cope with the ANFIS deficiencies, herein the PSO algorithm is coupled with the wave predictor FIS models in three viewpoints to optimize fuzzy subtractive clustering parameters, i.e. radii of clustering and quash factor, and the antecedent and consequent parameter of fuzzy IF-THEN rules. At first viewpoint, two PSO algorithms are used to optimize fuzzy subtractive clustering parameters and fuzzy IF-THEN rule parameters. In the second viewpoint, a PSO algorithm is used to optimize subtractive clustering parameters while the ANFIS model is used to tune the fuzzy IF-THEN rule parameters. In the third viewpoint, only a PSO algorithm is used to optimize the subtractive clustering parameters along with fuzzy IF-THEN rule parameters. Gathered data sets by National Data Buoy Center (NDBC) at Lake Michigan are used to evaluate the developed models for prediction of wave parameters including significant wave height and peak spectral period. Results indicate the efficiency of PSO algorithm to improve the ANFIS model accuracy.