Multi-Layer Hybrid Fuzzy Classification Based on SVM and Improved PSO for Speech Emotion Recognition

Speech Emotion Recognition (SER) plays a significant role in the field of Human–Computer Interaction (HCI) with a wide range of applications. However, there are still some issues in practical application. One of the issues is the difference between emotional expression amongst various individuals, and another is that some indistinguishable emotions may reduce the stability of the SER system. In this paper, we propose a multi-layer hybrid fuzzy support vector machine (MLHF-SVM) model, which includes three layers: feature extraction layer, pre-classification layer, and classification layer. The MLHF-SVM model solves the above-mentioned issues by fuzzy c-means (FCM) based on identification information of human and multi-layer SVM classifiers, respectively. In addition, to overcome the weakness that FCM tends to fall into local minima, an improved natural exponential inertia weight particle swarm optimization (IEPSO) algorithm is proposed and integrated with fuzzy c-means for optimization. Moreover, in the feature extraction layer, non-personalized features and personalized features are combined to improve accuracy. In order to verify the effectiveness of the proposed model, all emotions in three popular datasets are used for simulation. The results show that this model can effectively improve the success rate of classification and the maximum value of a single emotion recognition rate is 97.67% on the EmoDB dataset.

Fuzzy Classification Based Driving Distance Estimation for Electric Vehicles

Electric vehicle technology is an essential research field for improving full-electric vehicle (FEVs) capabilities. Different subsystem parameters in the FEVs should be monitored on a regular basis. The better these subsystems are used, the better the FEVs' performance, life, and range become. Nowadays, estimation of the state of charge (SoC) of the batteries and the driving distance is the area not been standardized sufficiently. In this study, a novel fuzzy classification method (FCM) is proposed to make the exact driving distance estimation of FEVs. The proposed FCM considers the consumed power and parameters of the battery under dynamic conditions. A test location was selected for the proposed FCM and tested under 3 different test conditions, namely, no-load, half-load and full-load conditions. Also, the performance of FCM is studied under several slope conditions, and the result shows that if the battery voltage decreases then the power consumed by the vehicle is improved in the uphill travel and the battery voltage is normal and the power consumption of the vehicle is decreased in the downhill drive. Finally, the drive distance of the proposed FCM is determined. HIGHLIGHTS Fuzzy classification based driving distance estimation for full-electric vehicle is proposed Parameters of battery and power consumption has been considered under dynamic condition CAN communication is established between different subsystems of electric vehicle Three test conditions (no-load, half load, and full load) have been considered

Acoustic Censusing and Individual Identification of Birds in the Wild

Avian vocal individuality carries information that can be utilized as an alternative to physical tagging of individuals. However, it is rarely used in conservation tasks despite rapidly-expanding use of passive acoustic monitoring techniques. Reliable acoustic individual recognizers and accurate quantifiers of population size remain elusive, which discourages the use of vocal individuality for monitoring, wildlife management, and ecological research. We propose a neuro-fuzzy framework that allows discrimination of individuals by their calls, the discovery of unexpected individuals in a set of recordings, and estimation of population size using solely sound. Our method, tested using data collected in the wild, allows rapid individual identification and even acoustic censusing without prior information from the recorded individuals. We achieve this by integrating a fuzzy classification and clustering methodology (LAMDA) into a Convolutional Deep Clustering Neural Network (CDCN). Our approach will benefit monitoring for conservation, and paves the way towards robust individual acoustic identification in species whose handling is time-consuming, culturally or ethically problematic, or logistically difficult.

Design of a Fuzzy Logic Classification System for an Ocular Sun Protection Factor

There is no reliable and universal method to assess and compare protective properties of lenses, which is generally communicated to the purchasers by sunglasses categories based on lenses only, classifying the entire sun transmission spectrum (ultraviolet, infrared and visible), without accounting the geometry of the frame, nor the back reflection of the lenses, which is a serious problem. The present work develops a fuzzy classification system whose purpose is to create a reliable method to rate and compare sunglasses through their sun protective properties. This system treats the pre-acquired data relative to the lenses, i.e. ultraviolet transmittance and back reflection, in addition to the frame coverage data through the fuzzification interface, followed by the fuzzy inference process and later by the defuzzification interface. The output is an alphabetic rate of the protective eyewear to inform consumers, in a simple way to understand, regarding the sun protective properties of the sunglasses they purchase. Through the results it was possible to prove the effectiveness of choosing a fuzzy system for this classification system, as it was able to translate, in mathematical terms, the linguistic rules, classifying each of the various possible combinations of the inputs to one of the six factors stipulated for the ocular sun protection factor (OSPF) proposed in this work.

Robust Algorithms for Change-Point Regressions Using the t-Distribution

Regression models with change-points have been widely applied in various fields. Most methodologies for change-point regressions assume Gaussian errors. For many real data having longer-than-normal tails or atypical observations, the use of normal errors may unduly affect the fit of change-point regression models. This paper proposes two robust algorithms called EMT and FCT for change-point regressions by incorporating the t-distribution with the expectation and maximization algorithm and the fuzzy classification procedure, respectively. For better resistance to high leverage outliers, we introduce a modified version of the proposed method, which fits the t change-point regression model to the data after moderately pruning high leverage points. The selection of the degrees of freedom is discussed. The robustness properties of the proposed methods are also analyzed and validated. Simulation studies show the effectiveness and resistance of the proposed methods against outliers and heavy-tailed distributions. Extensive experiments demonstrate the preference of the t-based approach over normal-based methods for better robustness and computational efficiency. EMT and FCT generally work well, and FCT always performs better for less biased estimates, especially in cases of data contamination. Real examples show the need and the practicability of the proposed method.

Algorithms and soft for adapting the knowledge base of project management information systems

The effectiveness of design solutions largely depends on the promptness of processing a large amount of data from various sources, which determines the feasibility of using information decision support systems (IDSS) in the field of project management. The peculiarities of information processes in project management greatly complicate or even make it impossible to implement in practice methods for constructing analytical, as well as probabilistic and statistical dependencies between the characteristics of the modeled project management system and the indicators of its internal and external environment. In this regard, as an algorithmic support for IDSS for project management, it is promising to use precedent methods for analyzing information based on knowledge about similar situations previously observed in the practice of project management, and representing knowledge in the form of ontologies. Analysis of practical situations in the field of project management makes it possible to substantiate the expediency of organizing a monitoring procedure for the IDSS knowledge base, based on the results of which decisions on its adaptation are made. The article proposes the main ways of this adaptation: changing the structure and basic elements (first of all, concepts) of ontologies; clarification of the structure of the description of current situations and, therefore, precedents. The developed algorithm for monitoring the IDSS knowledge base on project management for the analysis and identification of typical situations of the feasibility of changing it is described. The algorithm is distinguished by the possibility of developing recommendations on the modification of ontologies based on a fuzzy classification of search results and using precedents relevant to current situations. A procedure is proposed for changing the structure of the description of precedents, taking into account the results of assessing the indices of the fuzzy correspondence of the characteristics of the existing precedents to the characteristics of the project being implemented. A description of a computer program that implements the proposed algorithm and its components, as well as the results of its application are given.

Fuzzy Evaluation of Kansei Attributes Using Convolutional Neural Networks

The emotional needs for products have increased significantly with the recent improvements in living standards. Attribute evaluation forms the core of Kansei engineering in emotion-oriented products, and is practically quite subjective in nature. Essentially, attribute evaluation is a fuzzy classification task, whose quantitative results change slightly with statistical time and statistical objects, making it difficult to accurately describe using standard mathematical models. In this paper, we propose a novel deep-learning-assisted fuzzy attribute-evaluation (DLFAE) method, which could generate quantitative evaluation results. In comparison to existing methods, the proposed method combines subjective evaluation with convolutional neural networks, which facilitates the generation of quantitative evaluation results. Additionally, this strategy has better transferability for different situations, increasing its versatility and applicability. This, in turn, reduces the computational burden of evaluation and improves operational efficiency.