Exploring Sequential Feature Selection in Deep Bi-LSTM Models for Speech Emotion Recognition

Machine Learning(ML)algorithms play a pivotal role in Speech Emotion Recognition(SER),although they encounter a formidable obstacle in accurately discerning a speaker’s emotional state.The examination of the emotional states of speakers holds significant importance in a range of real-time applications,including but not limited to virtual reality,human-robot interaction,emergency centers,and human behavior assessment.Accurately identifying emotions in the SER process relies on extracting relevant information from audio inputs.Previous studies on SER have predominantly utilized short-time characteristics such as Mel Frequency Cepstral Coefficients(MFCCs)due to their ability to capture the periodic nature of audio signals effectively.Although these traits may improve their ability to perceive and interpret emotional depictions appropriately,MFCCS has some limitations.So this study aims to tackle the aforementioned issue by systematically picking multiple audio cues,enhancing the classifier model’s efficacy in accurately discerning human emotions.The utilized dataset is taken from the EMO-DB database,preprocessing input speech is done using a 2D Convolution Neural Network(CNN)involves applying convolutional operations to spectrograms as they afford a visual representation of the way the audio signal frequency content changes over time.The next step is the spectrogram data normalization which is crucial for Neural Network(NN)training as it aids in faster convergence.Then the five auditory features MFCCs,Chroma,Mel-Spectrogram,Contrast,and Tonnetz are extracted from the spectrogram sequentially.The attitude of feature selection is to retain only dominant features by excluding the irrelevant ones.In this paper,the Sequential Forward Selection(SFS)and Sequential Backward Selection(SBS)techniques were employed for multiple audio cues features selection.Finally,the feature sets composed from the hybrid feature extraction methods are fed into the deep Bidirectional Long Short Term Memory(Bi-LSTM)network to discern emotions.Since the deep Bi-LSTM can hierarchically learn complex features and increases model capacity by achieving more robust temporal modeling,it is more effective than a shallow Bi-LSTM in capturing the intricate tones of emotional content existent in speech signals.The effectiveness and resilience of the proposed SER model were evaluated by experiments,comparing it to state-of-the-art SER techniques.The results indicated that the model achieved accuracy rates of 90.92%,93%,and 92%over the Ryerson Audio-Visual Database of Emotional Speech and Song(RAVDESS),Berlin Database of Emotional Speech(EMO-DB),and The Interactive Emotional Dyadic Motion Capture(IEMOCAP)datasets,respectively.These findings signify a prominent enhancement in the ability to emotional depictions identification in speech,showcasing the potential of the proposed model in advancing the SER field.

Selection Method of Multi-Objective Problems Using Genetic Algorithm in Motion Plan of AUV

To research the effect of the selection method of multi — objects genetic algorithm problem on optimizing result, this method is analyzed theoretically and discussed by using an autonomous underwater vehicle (AUV) as an object. A changing weight value method is put forward and a selection formula is modified. Some experiments were implemented on an AUV, TwinBurger. The results shows that this method is effective and feasible.

Feature Selection for Multi-label Classification Using Neighborhood Preservation

Multi-label learning deals with data associated with a set of labels simultaneously. Dimensionality reduction is an important but challenging task in multi-label learning. Feature selection is an efficient technique for dimensionality reduction to search an optimal feature subset preserving the most relevant information. In this paper, we propose an effective feature evaluation criterion for multi-label feature selection, called neighborhood relationship preserving score. This criterion is inspired by similarity preservation, which is widely used in single-label feature selection. It evaluates each feature subset by measuring its capability in preserving neighborhood relationship among samples. Unlike similarity preservation, we address the order of sample similarities which can well express the neighborhood relationship among samples, not just the pairwise sample similarity. With this criterion, we also design one ranking algorithm and one greedy algorithm for feature selection problem. The proposed algorithms are validated in six publicly available data sets from machine learning repository. Experimental results demonstrate their superiorities over the compared state-of-the-art methods.

Fuzzy Multi-Objective Decision Model of Supplier Selection with Preference Information

Supplier selection is a multi-objective decision problem, which must be considered many objectives, some objectives are qualitative, and others are quantitative. Meanwhile, manufacturer has preference for different suppliers. In this paper, a new multi-objective decision model with preference information of supplier is established. A practical example of supplier selection problem utilizing this model is studied. The result demonstrates the feasibility and effectiveness of the methods proposed in the paper.

Phase selection rules for complex multi-component alloys with equiatomic or close-to-equiatomic compositions

Alloying greatly expands the amount of available materials beyond the naturally existing ones, and more importantly offers the material scientists opportunities to initiatively control the composition-structure-property relationship in materials. Since commonly used metallic materials are mostly multi-component alloys, the know-how of alloying through compositional control, certainly plays a critical role in designing materials with desired structure and properties. However, alloying in multi-component alloys is an extremely complicated issue, as the alloyed products could be the amorphous phase, various solid solutions and intermetallic compounds containing two or more alloy components. By narrowing down the scope of the multi-component alloys to those with equiatomic or close-to-equiatomic compositions only, and also aiming at framing out the rules that govern the phase selection upon alloying in multi-component alloys in a broad sense, we have identified here a simple and easily executable two-parameter scheme that can effectively predict the formation of the amorphous phase, solid solutions and intermetallic compounds, in multi-component alloys, simply from the given alloy compositions. We believe this scheme reveals a clear physical scenario governing the phase selection in multi-component alloys, helps to simplify the alloy design, and benefits the future development of advanced metallic alloys like bulk metallic glasses and high entropy alloys.

Do Search and Selection Operators Play Important Roles in Multi-Objective Evolutionary Algorithms:A Case Study

Multi-objective Evolutionary Algorithm (MOEA) is becoming a hot research area and quite a few aspects of MOEAs have been studied and discussed. However there are still few literatures discussing the roles of search and selection operators in MOEAs. This paper studied their roles by solving a case of discrete Multi-objective Optimization Problem (MOP): Multi-objective TSP with a new MOEA. In the new MOEA, We adopt an efficient search operator, which has the properties of both crossover and mutation, to generate the new individuals and chose two selection operators: Family Competition and Population Competition with probabilities to realize selection. The simulation experiments showed that this new MOEA could get good uniform solutions representing the Pareto Front and outperformed SPEA in almost every simulation run on this problem. Furthermore, we analyzed its convergence property using finite Markov chain and proved that it could converge to Pareto Front with probability 1. We also find that the convergence property of MOEAs has much relationship with search and selection operators.

Zero-forcing beamforming with receiver antenna selection in downlink multi-antenna multi-user system

A study on the zero-forcing beamforming （ZFBF） scheme with antenna selection at user terminals in downlink multi-antenna multi-user systems is presented. Simulation results show that the proposed ZFBF scheme with receiver antenna selection （ZFBF-AS） achieves considerable throughput improvement over the ZFBF scheme with single receiver antenna. The results also show that, with multi-user diversity, the ZFBF-AS scheme approaches the throughput performance of the ZFBF scheme using all receiver antennas （ZFBF-WO-AS） when the base station adopts semi-orthogonal user selection （SUS） algorithm, and achieves larger throughput when the base station adopts the Round-robin scheduling algorithm. Compared with ZFBF-WO-AS, the proposed ZFBF-AS scheme can reduce the cost of user equipments and the channel state information requirement at the transmitter （CSIT） as well as the multiuser scheduling complexity at the transmitter.

Algorithm for multi-constrained path selection based on experimental analysis

It is a challenging problem to provide quality-of-service （QoS） guarantees in next generation high-speed network, and the QoS routing is one of the key issues of the problem. For the problem of multi-constrained QoS routing in high-speed network, especially under the inaccurate link state information, the success ratio of the different constraint combination is analyzed statistically, and a constraint analysis method based on the computer simulation is proposed. Furthermore, the approximately equal loose-tight order relation between each two constraints is constructed, and then an algorithm based on the experimental analysis is presented. Finally, the simulation result demonstrates that the algorithm has the higher success ratio, and the theoretical analysis proves its correctness and universality.

Path Selection of Multimodal Transport Based on Multi-Objective Mixed Integer Programming

Based on “One Belt and One Road”, this paper studies the path selection of multimodal transport by using the method of multi-objective mixed integer programming. Therefore, this paper studies the factors of transportation time, transportation cost and transportation safety performance, and establishes a mathematical model. In addition, the method of multi-objective mixed integer programming is used to comprehensively consider the different emphasis and differences of customers on cargo transportation. Then we use planning tools of Microsoft Excel to solve path selection and to determine whether the chosen path is economical and reliable. Finally, a relatively complex road network is built as an example to verify the accuracy of this planning method.

A Multiobjective Optimization Algorithm for QoS-Aware Path Selection in DiffServ and MPLS Networks

A multiobjective quality of service (QoS) routing algorithm was proposed and used as the QoS-aware path selection approach in differentiated services and multi-protocol label switching (DiffServ-MPLS) networks. It simultaneously optimizes multiple QoS objectives by a genetic algorithm in conjunction with concept of Pareto dominance. The simulation demonstrates that the proposed algorithm is capable of discovering a set of QoS-based near optimal paths within in a few iterations. In addition, the simulation results also show the scalability of the algorithm with increasing number of network nodes.

An efficient multi-relay selection scheme in cooperative communication system

Cooperative relaying has played an important role in rapid evolution of wireless communications.The cooperative performance strongly depends on the selected relays.In this paper,we concentrate on relay selection in amplify-and-forward(AF)cooperative communication system,and an optimal multi-relay selection scheme is put forward to minimize the average symbol error rate(SER)of the system.Firstly,for the minimum average SER,on the basis of the statistic channel information,we define a parameter named equivalent channel gain which describes the channel status of two phases in the cooperative process.Then,under the constraint of equal power allocation,an optimal relay selection scheme is proposed in ascending order of equivalent channel gain(ECG).The scheme implies that the suitable number of relay nodes should be selected under the different signal-to-noise ratio(SNR)ranges to minimize the average SER.Computer simulation results show that the average SER rate of the proposed scheme is lower than these of the other schemes.

PARAMETRIC MATCHING SELECTION OF MULTI-MEDIUM COUPLING SHOCK ABSORBER

To achieve the dual demand of resisting violent impact and attenuating vibration in vibration-impact-safety of protection for precision equipment such as MEMS packaging system, a theo- retical mathematical model of multi-medium coupling shock absorber is presented. The coupling of quadratic damping, linear damping, Coulomb damping and nonlinear spring are considered in the model. The approximate theoretical calculating formulae are deduced by introducing transformation-tactics. The contrasts between the analytical results and numerical integration results are developed. The resisting impact characteristics of the model are also analyzed in progress. In the meantime, the optimum model of the parameters matching selection for design of the shock absorber is built. The example design is illustrated to confirm the validity of the modeling method and the theoretical solution.

EXPERIMENTAL STUDIES OF MEDICAL EQUIPMENT SELECTION BY USING THE MULTI-STAGE WEIGHTING EVALUATION METHOD

As more and more sophiscated medical equipment is being purchased byhospitals, selection is becoming an increasinly complex and important process. But it'sdifficult to compare different products analytically. Therefore personal preferences andbiases will probably be ntroduced. This paper describes a scientific model selectionmethod to quantify a product's value to make comparison easy.

A Multi-Agent Resource Selection Model in Manufacturing Grid Environment

A group of agents are intimately cooperated to set the assessment indices, establish the weight of each index in overall result of evaluation, collect the experts＇ scores given to each available resource, and the manufacturing resource whose overall assessment value is highest is taken as the optimal choice. Architecture of the proposed system is outlined and an example is offered to show the process of accomplishing the assessment.

Optimal Test Points Selection Based on Multi-Objective Genetic Algorithm

A new approach to select anoptimal set of test points is proposed. The described method uses fault-wise table and multi-objective genetic algorithm to find the optimal set of test points. First, the fault-wise table is constructed whose entries are measurements associated with faults and test points. The selection of optimal test points is transformed to the selection of the columns that isolate the rows of the table. Then, four objectives are described according to practical test requirements. The multi-objective genetic algorithm is explained. Finally, the presented approach is illustrated by a practical example. The results indicate that the proposed method can efficiently and accurately find the optimal set of test points and is practical for large scale systems.

A Multi-Hop Dynamic Path-Selection (MHDP) Algorithm for the Augmented Lifetime of Wireless Sensor Networks

In Wireless Sensor Networks (WSN), the lifetime of sensors is the crucial issue. Numerous schemes are proposed to augment the life time of sensors based on the wide range of parameters. In majority of the cases, the center of attraction will be the nodes’ lifetime enhancement and routing. In the scenario of cluster based WSN, multi-hop mode of communication reduces the communication cast by increasing average delay and also increases the routing overhead. In this proposed scheme, two ideas are introduced to overcome the delay and routing overhead. To achieve the higher degree in the lifetime of the nodes, the residual energy (remaining energy) of the nodes for multi-hop node choice is taken into consideration first. Then the modification in the routing protocol is evolved (Multi-Hop Dynamic Path-Selection Algorithm—MHDP). A dynamic path updating is initiated in frequent interval based on nodes residual energy to avoid the data loss due to path extrication and also to avoid the early dying of nodes due to elevation of data forwarding. The proposed method improves network’s lifetime significantly. The diminution in the average delay and increment in the lifetime of network are also accomplished. The MHDP offers 50% delay lesser than clustering. The average residual energy is 20% higher than clustering and 10% higher than multi-hop clustering. The proposed method improves network lifetime by 40% than clustering and 30% than multi-hop clustering which is considerably much better than the preceding methods.

Design of multi-band frequency selective surfaces using multi-periodicity combined elements

Traditional multi-band frequency selective surface （FSS） approaches are hard to achieve a perfect resonance response in a wide band due to the limit of the onset grating lobe frequency determined by the array. To solve this problem, an approach of combining elements in different period to build a hybrid array is presented. The results of series of numerical simulation show that multi-periodicity combined element FSS, which are designed using this approach, usually have much weaker grating lobes than the traditional FSS. Furthermore, their frequency response can be well predicted through the properties of their member element FSS. A prediction method for estimating the degree of expected grating lobe energy loss in designing multi-band FSS using this approach is provided.

Genetic algorithm for pareto optimum-based route selection

A quality of service （QoS） or constraint-based routing selection needs to find a path subject to multiple constraints through a network. The problem of finding such a path is known as the multi-constrained path （MCP） problem, and has been proven to be NP-complete that cannot be exactly solved in a polynomial time. The NPC problem is converted into a multiobjective optimization problem with constraints to be solved with a genetic algorithm. Based on the Pareto optimum, a constrained routing computation method is proposed to generate a set of nondominated optimal routes with the genetic algorithm mechanism. The convergence and time complexity of the novel algorithm is analyzed. Experimental results show that multiobjective evolution is highly responsive and competent for the Pareto optimum-based route selection. When this method is applied to a MPLS and metropolitan-area network, it will be capable of optimizing the transmission performance.

Strategic Renewable Energy Resource Selection Using a Fuzzy Decision-Making Method

Renewable energy is created by renewable natural resources such as geothermal heat,sunlight,tides,rain,and wind.Energy resources are vital for all countries in terms of their economies and politics.As a result,selecting the optimal option for any country is critical in terms of energy investments.Every country is nowadays planning to increase the share of renewable energy in their universal energy sources as a result of global warming.In the present work,the authors suggest fuzzy multi-characteristic decision-making approaches for renew-able energy source selection,and fuzzy set theory is a valuable methodology for dealing with uncertainty in the presence of incomplete or ambiguous data.This study employed a hybrid method for order of preference by resemblance to an ideal solution based on fuzzy analytical network process-technique,which agrees with professional assessment scores to be linguistic phrases,fuzzy numbers,or crisp numbers.The hybrid methodology is based on fuzzy set ideologies,which calculate alternatives in accordance with professional functional requirements using objective or subjective characteristics.The best-suited renewable energy alternative is discovered using the approach presented.