Superior neuro-fuzzy classification systems

Position and Tilt Control of Two-Wheeled Robot (TWR)

This paper presents a fuzzy based adaptive control approach for stabilization of Two wheeled robot (TWR) system. The TWR consists of a robot chassis mounted on two movable wheels. The objective is to stabilize the proposed system within desired time, minimum overshoot and at desired location. The data samples collected from simulation results of fuzzy controllers were used for training, tuning and optimisation of an adaptive neuro fuzzy inference system(ANFIS) controller. A Matlab Simulink model of the system has been built using Newton's second law of motion. The effect of shape and number of membership functions on training error of ANFIS has also been analysed. The designing of fuzzy rules for both fuzzy and ANFIS controller were carried out using gbell shape memberships. Simulations were performed which compared and validated the performance of both the controllers.

Hierarchal Fuzzy Logic Controller and Internet of Things (IoT) Information

With the advances of networks and sensing technologies, it is possible to benefit from the surrounding environment's data in enhancing peoples' life. Currently, we have different types of networks such as Wireless Sensor Networks (WSNs), Vehicle Ad Hoc Networks (VANETs), Cellular Networks (CNs), and Social Networks (SNs) along with underlying computing such as Cloud computing. These types of networks provide huge data about the surrounding environments including weather information, peoples' relations, peoples' interest, and location information. This paper examines the suitability of hierarchal fuzzy logic controller in classifying the IoT data. The paper also tries to answer “if-else “questions about the effect of each of the input parameters. The authors' test case in this paper is related to the disease spreading prediction problem. This test case is highly important to the health care organizations. Different case studies are generated to examine the efficiency of the proposed framework and methodologies.

Breast Cancer Detection with Reduced Feature Set

This paper explores feature reduction properties of independent component analysis (ICA) on breast cancer decision support system. Wisconsin diagnostic breast cancer (WDBC) dataset is reduced to one-dimensional feature vector computing an independent component (IC). The original data with 30 features and reduced one feature (IC) are used to evaluate diagnostic accuracy of the classifiers such as k-nearest neighbor (k-NN), artificial neural network (ANN), radial basis function neural network (RBFNN), and support vector machine (SVM). The comparison of the proposed classification using the IC with original feature set is also tested on different validation (5/10-fold cross-validations) and partitioning (20%–40%) methods. These classifiers are evaluated how to effectively categorize tumors as benign and malignant in terms of specificity, sensitivity, accuracy, F-score, Youden's index, discriminant power, and the receiver operating characteristic (ROC) curve with its criterion values including area under curve (AUC) and 95% confidential interval (CI). This represents an improvement in diagnostic decision support system, while reducing computational complexity.