Search

Many real-world situations such as bad weather may result in hazy environments. Images captured in these hazy conditions will have low image quality due to microparticles in the air. The microparticles light to scatter and absorb, resulting in hazy images with various effects. In recent years, image dehazing has been researched in depth to handle images captured in these conditions. Various methods were developed, from traditional methods to deep learning methods. Traditional methods focus more on the use of statistical prior. These statistical prior have weaknesses in certain conditions. This paper proposes a novel architecture based on PDR-Net by using a pyramid dilated convolution and pre-processing modules, processing modules, post-processing modules, and attention applications. The proposed network is trained to minimize L1 loss and perceptual loss with the O-Haze dataset. To evaluate our architecture's result, we used structural similarity index measure (SSIM), peak signal-to-noise ratio (PSNR), and color difference as an objective assessment and psychovisual experiment as a subjective assessment. Our architecture obtained better results than the previous method using the O-Haze dataset with an SSIM of 0.798, a PSNR of 25.39, but not better on the color difference. The SSIM and PSNR results were strengthened by using subjective assessments and 65 respondents, most of whom chose the results of the restoration of the image produced by our architecture.

The Data Envelopment Analysis (DEA) method is a method commonly used in benchmarking. The Dynamic Data Envelopment Analysis (DDEA) method was proposed to improve the DEA method in the benchmarking process. The DDEA method proposed can determine the effectiveness of the Decision Making Unit (DMU). The disadvantage of the DDEA model is that it cannot handle problems that involve benchmarking for stochastic data. To improve the DDEA method, the Stochastic Data Envelopment Analysis (SDEA) method is proposed which can be used for benchmarking involving stochastic data. The SDEA method itself has weaknesses in dealing with noise and uncertainty problems that will appear in the assessment process. The purpose of the research conducted by the researcher was to use the Hesitant Fuzzy method in optimizing the SDEA method so that the Hesitant Fuzzy model - Stochastic Data Envelopment Analysis (HF-SDEA) could be carried out benchmarking process in a situation where the assessment contained many elements of uncertainty. The results of this study are benchmarking methods that can do benchmarking for stochastic data on conditions that contain elements of uncertainty.

A company maintains and improves its quality services by paying attention to reviews and complaints from users. The complaints from users are commonly written using human natural language expression so that their messages are computationally difficult to extract and proceed. To overcome this difficulty, in this study, we presented a new system for issues feature extraction from users’ reviews and complaints from social media data. This system consists of four main functions: (1) Data Crawling and Preprocessing, (2) Categorization Knowledge Modelling, (3) Rule-based Sentiment Analysis, and (4) Application Environment. Data Crawling and Preprocessing provides data acquisition from users’ tweets on social media, crawls the data and applies the data preprocessing. Categorization Knowledge Modelling provides text mining of textual data, vector space transformation to create knowledge metadata, context recognition of keyword queries to the knowledge metadata, and similarity measurement for categorization. In the Rule-based Sentiment Analysis, we developed our own rules of computatioal linguistics to measure polarity of sentiment. Application Environment consists of 3 layers: database management, back-end services and front-end services. For applicability of our proposed system, we conducted two kinds of experimental study: (1) categorization performance, and (2) sentiment analysis performance. For categorization performance, we used 8743 tweet data and performed 82% of accuracy. For categorization performance, we made experiments on 217 tweet data and performed 92% of accuracy.

Entrepreneurship plays an essential role in the economic growth of a country. These roles include creating jobs, reducing unemployment, increasing people's income, combining production factors (nature, labor, capital, and expertise), and increasing national productivity. For the economy to thrive and healthy, it requires at least 4% of the population who work as entrepreneurs. Due to this vital role, entrepreneurial growth must be maintained. One of the efforts to do this is by monitoring growth directly and continuously. Besides that, another way is to do a simulation. By knowing the condition of entrepreneurship at one time and all the factors that affect entrepreneurial growth, simulations can be carried out to determine or predict future conditions. Based on this simulation, essential steps can be taken, or policies can be made to maintain profitable entrepreneurial growth. This paper presents a mathematical model that can simulate and visualize entrepreneurship's growth in six provinces of Sumatra Island, Indonesia. This mathematical model uses cellular automata as its basis and is called Entrepreneurial Cellular Automata (ECA). One of the advantages of Cellular Automata is that it is easy to visualize. The entrepreneurial model used as a reference is a model from the Global Entrepreneurship Monitoring (GEM). This mathematical model has been implemented in a simulator program. This paper describes the simulator development and the use of simulator to simulate and visualize the entrepreneurial growth of the six provinces.

Smart City Mobile Apps is currently being developed by property developers in major cities in Indonesia and is used by businesses to market their products through the application. Technology integration in city governance is possible, thanks to the internet of things, a network of electronic devices interconnected and capable of sending data or following up with minimal human intervention. This study aims to identify the characteristics of Smart City Mobile Apps users in the South Tangerang area. The method used to see how well the implementation of digital marketing in Smart City Mobile Apps combines the TAM (Technology Acceptance Model) and UTAUT (Unified Theory of Acceptance and Use of Technology) models. Questionnaire data were processed using the Structural Equation Modelling (SEM) method. The study results state that the company's digital marketing strategy through Smart City Mobile Apps is already satisfying. It can be seen from the value of adoption of Smart City Mobile Apps users in utilizing this application, which is considerably large at 0.764 based on the coefficient of determination. The attitude variable influences consumers to use the application in finding trade information in the Smart City, with an influence value of 0.412 and the behavior intention to use with an influence value of 0.726, which shows that the intended behavior of the application and the trust in the benefits of the application can encourage users to use this application for purchasing their daily needs.

Agent based simulation (ABS) is a paradigm to modelling systems included of autonomous and interacting agents. ABS has been tremendous growth and used by researchers in the social sciences to study socio-environmental complex systems. To date, various platforms have been introduced for agent-based social simulation. They are rule based in any logic, python based in SPADE and etc. Although those platforms have been introduced, there is still an insufficient to develop a crowd simulation in 3D platform. Having a 3D platform is needed to enabling the crowd simulation for training purposes. However, the current tools and platform still lack features to develop and simulate autonomous agents in the 3D world. This paper introduced a BDI plug in at Unity3D for crowd simulation. BDI is an intelligent agent architecture and it is able to develop autonomous agents in crowd environment. In this paper, we present the BDI plug with a case study of Australia bush fire and discuss a method to support autonomous agents' development in 3D crowd simulation. The tool allows the modeller to develop autonomous agents in 3D world by taking the advantages of Unity3D.

Class imbalance is one of the main problems in classification because the number of samples in majority class is far more than the number of samples in minority class. The class imbalance problem in the multi-class dataset is much more difficult to handle than the problem in the two class dataset. This multi-class imbalance problem is even more complicated if it is accompanied by overlapping. One method that has proven reliable in dealing with this problem is the Hybrid Approach Redefinition-Multiclass Imbalance (HAR-MI) method which is classified as a hybrid approach which combines sampling and classifier ensembles. However, in terms of diversity among classifiers, hybrid approach that combine sampling and classifier ensembles will give better results. HAR-MI delivers excellent results in handling multi-class imbalances. The HAR-MI method uses SMOTE to increase the number of sample in minority class. However, this SMOTE also has a weakness where if there is an extremely imbalanced dataset and a large number of attributes there will be over-fitting. To overcome the problem of over-fitting, the Hybrid Sampling method was proposed. HAR-MI combination with Hybrid Sampling is done to increase the number of samples in the minority class and at the same time reduce the number of noise samples in the majority class. The preprocessing stages at HAR-MI will use the Minimizing Overlapping Selection under Hybrid Sazmpling (MOSHS) method and the processing stages will use Different Contribution Sampling. The results obtained will be compared with the results using Neighbourhood-based undersampling. Overlapping and Classifier Performance will be measured using Augmented R-Value, the Matthews Correlation Coefficient (MCC), Precision, Recall, and F-Value. The results showed that HAR-MI with Hybrid Sampling gave better results in terms of Augmented R-Value, Precision, Recall, and F-Value.

MDM is a trade cooperative business unit that supplies healthy food options for consumers around the Andalas University campus. So far, the selection of suppliers that provide supply goods to the stores is only based on the trust between both parties, which is the principle of mutual acquaintance and kinship. The problems that may arise from a process like this are the lack of the right supplier, unavailability of goods, relatively higher product prices, late delivery, and low-quality goods. Therefore, we need a GDSS that is capable of overcoming these problems. This GDSS helps in decision-making by determining the right supplier for each of the stores owned by the MDM Cooperative. The methods used are AHP, TOPSIS, and BORDA, involving six criteria and five tested alternatives. The AHP method is used to obtain the weight of each criterion taken from the pairwise comparison matrix. The TOPSIS method is used to determine which suppliers get priority for supply goods. Combining the AHP and TOPSIS methods can reduce the weaknesses of the TOPSIS itself by giving subjective weights. The use of the BORDA method can provide maximum results in selecting this supplier. This GDSS also involves three decision-making bodies: the head of the cooperative, the deputy, and the treasurer. The results of this prototype can show the best alternative selected based on the ranking method.

The impact of this pandemic affects various sectors in Indonesia, especially in the economic sector, due to the large-scale social restrictions policy to suppress this case's growth. The details of the growth of Covid-19 in Indonesia are still fluctuating and cannot be fully understood. Recently it has been developed by researchers related to the prediction of Covid-19 cases in various countries. One of them is using a machine learning technique approach to predict cases of daily increase Covid-19. However, the use of machine learning techniques results in the MSE error value in the thousands. This high number indicates that the prediction data using the model is still a high error rate compared to the actual data. In this study, we propose a deep learning approach using the Long Short Term Memory (LSTM) method to build a prediction model for the daily increase cases of Covid-19. This study's LSTM model architecture uses the LSTM layer, Dropout layer, Dense, and Linear Activation Function. Based on various hyperparameter experiments, using the number of neurons 10, batch size 32, and epochs 50, the MSE values were 0.0308, RMSE 0.1758, and MAE 0.13. These results prove that the deep learning approach produces a smaller error value than machine learning techniques, even closer to zero.

Natural language processing (NLP) is a branch of computer science concerned with the understanding of human language and communication, and translating these into a computer-comprehensible embedding. Our goal in this paper is to capture meaning from human natural language through NLP and provide an automated solution for aiding the process of service ticket solving, through the intelligent classification of tickets, pattern recognition and similarities between texts. The difficulty of this task lies in translating the human language into a mathematical format: transforming a non-formal language, into a formal one, without losing any details. Also what raises even more complication is the context in which this language appears: service tickets, that come from a technical and specialized jargon of computer science and IT industry, and the brief manner in which the tickets are written. This paper aims to tackle this challenge through multiple methods of text classification and recognition, and data analysis, followed by comparison and interpretation of the results. In completion, we find that our methods yield plausible results to be implemented in helping the service process.

This paper investigates the performance of a class of irregular low-density parity-check (LDPC) codes through a recently published low complexity upper bound on their belief propagation decoding thresholds. Moreover, their performance analysis is carried out through a recently published algorithmic method, presented in Babich et al. 2017 paper. In particular, the class considered is characterized by variable node degree distributions (lambda(x)) of minimum degree (i_1 gt 2): being, in this case, (lambda^{'} (0)=lambda_2=0), this is useful to design LDPC codes presenting a linear minimum distance growth with the block length with probability 1, as shown in Di et al.’s 2006 paper. These codes unfortunately cannot reach capacity under iterative decoding, since the achievement of capacity requires (lambda_2 neq 0). However, in this latter case, the block error probability might converge to a constant, as shown in the aforementioned paper.

In a passive radar system, localizing a target in Cartesian space is achieved by using one of the following bistatic geometries: multiple non-cooperative transmitters with one receiver, one non-cooperative transmitter with multiple receivers, or one non-cooperative transmitter with one receiver. In this paper, we propose a new method for localizing a target in Cartesian space by passive radar having the bistatic geometry “one non-cooperative transmitter and one receiver”. This method depends on using two consecutive particle filters for estimating and analyzing the Doppler frequency and time delay of the target’s echo signal. The theoretical analysis of the proposed method is presented, and its efficiency is verified by simulating the passive radar system with a Digital Video Broadcasting-Terrestrial (DVB-T) transmitter.

Mobile ad-hoc network is a collection of dynamically organized nodes where each node acts as a host and router. Mobile ad-hoc networks are characterized by the lack of preexisting infrastructures or centralized administration. So, they are vulnerable to several types of attacks, especially the Blackhole attack. This attack is one of the most serious attacks in this kind of mobile networks. In this type of attack, the malicious node sends a false answer indicating that it has the shortest path to the destination node by increasing the sequence number and decreasing the number of hops. This will have a significant negative impact on source nodes which send their data packets through the malicious node to the destination. This malicious node drops received data packets and absorbs all network traffic. In order overcome this problem, securing routing protocols become a very important requirement in mobile ad-hoc networks. Multi-path routing protocols are among the protocols affected by the Blackhole attack. In this paper, we propose an effective and efficient technique that avoids misbehavior of Blackhole nodes and facilitates the discovery for the most reliable paths for the secure transmission of data packets between communicating nodes in the well known Ad hoc On-demand multi-path routing protocol (AOMDV). Our proposed technique is implemented and simulated using the well known ns 2.35 simulator. We also compared the performance of the three routing protocols AOMDV, AOMDV under Blackhole attack (BHAOMDV) and the proposed solution to counter the Blackhole attack (IDSAOMDV). The results show the degradation of performance of AOMDV under attack, it also present similarities between normal AOMDV and the proposed solution by isolating misbehaving node which has resulted in increase the performance metrics to the standard values of the AOMDV protocol.

Software companies usually develop a set of product variants within the same family that share certain functions and differ in others. Variations across software variants occur to meet different customer requirements. Thus, software product variants evolve overtime to cope with new requirements. A software engineer who deals with this family may find it difficult to understand the evolution scenarios that have taken place over time. In addition, software identifier names are important resources to understand the evolution scenarios in this family. This paper introduces an automatic approach called Juana's approach to detect the evolution scenario across two product variants at the source code level and identifies the common and unique software identifier names across software variants source code. Juana's approach refers to common and unique identifier names as a software identifiers map and computes it by comparing software variants to each other. Juana considers all software identifier names such as package, class, attribute, and method. The novelty of this approach is that it exploits common and unique identifier names across the source code of software variants, to understand the evolution scenarios across software family in an efficient way. For validity, Juana was applied on ArgoUML and Mobile Media software variants. The results of this evaluation validate the relevance and the performance of the approach as all evolution scenarios were correctly detected via a software identifiers map., Comment: 9 pages, 7 figures, 6 tables

The purpose of the article is to study and consider general trends, problems and prospects for the use of ontological modelling of investment activity. Research methods are methods of semantic analysis of the basic concepts of the considered subject area (investment activity). The article discusses approaches to ontological modelling of investment activities and the development of modern intelligent expert systems that support investment processes. The novelty of the research is the analysis of the use of ontological modelling in the optimization of investment activities The conclusion of the research conducted in the article is as follows: the article considers various aspects related to intelligent technologies based on modeling, in particular on the ontological; the essence of ontological modelling of investment activity is considered. Use of cognitive-ontological model promotes the formation of optimum administrative decisions in the field of investment activity. The set of interrelated management actions for investment activities, their logical sequence constitute a holistic management strategy – a model of process management in the system of investment activities.

Data traffic demand over the Internet is increasing rapidly, and it is changing the pricing model between internet service providers (ISPs), content providers (CPs) and end users. One recent pricing proposal is sponsored data plan, i.e., when CP negotiates with the ISP on behalf of the users to remove the network subscription fees so as to attract more users and increase the number of advertisements. As such, a key challenge is how to provide proper sponsorship in the situation of complex interactions among the telecommunication actors, namely, the advertisers, the content provider, and users. To answer those questions, we explore the potential economic impacts of this new pricing model by modeling the interplay among the advertiser, users, and the CPs in a game theoretic framework. The CP may have either a subscription revenue model (charging end-users) or an advertisement revenue model (charging advertisers). In this work, we design and analyze the interaction among CPs having an advertisement revenue as a non-cooperative game, where each CP determines the proportion of data to sponsor and a level of credibility of content. In turn, the end-users demand for the content of a CP depends not only on their strategies but also upon those proposed by all of its competitors. Through rigorous mathematical analysis, we prove the existence and uniqueness of the Nash equilibrium. Based on the analysis of the game properties, we propose an iterative algorithm, which guarantees to converge to the Nash equilibrium point in a distributed manner. Numerical investigation shows the convergence of a proposed algorithm to the Nash equilibrium point and corroborates the fact that sponsoring content may improve the CPs outcome.

In a cognitive radio network (CRN), the primary users (PUs) do not operate their spectra, full time. Thus, they can sell them to the secondary users (SUs), for a second use, during the free time slots. In this article, we assume that the market is perfect, monopolized by a single PU, and all players are rational. After formulating the PU’s profit, we established a necessary and sufficient condition that guarantees the introduction of the PU into the market. In addition, the expressions of the SUs’ profits, showed us that in non-cooperative form, some ones got zero profit, even after maximizing their profits. Therefore, we have considered to study the effect of cooperation on the profits of this category of SUs. By following this step, we established a cooperation strategy, to avoid zero profits for all SUs. In order to analyze the impact of this cooperation on the PU, we have expressed the profits of the PU in the cooperative and non-cooperative forms; as result, we found that the cooperation between SUs brought better than the non-cooperative form.

Much attention has been paid to large data technologies in the past few years mainly due to its capability to impact business analytics and data mining practices, as well as the possibility of influencing an ambit of a highly effective decision-making tools. With the current increase in the number of modern applications (including social media and other web-based and healthcare applications) which generates high data in different forms and volume, the processing of such huge data volume is becoming a challenge with the conventional data processing tools. This has resulted in the emergence of big data analytics which also comes with many challenges. This paper introduced the use of principal components analysis (PCA) for data size reduction, followed by SVM parallelization. The proposed scheme in this study was executed on the Spark platform and the experimental findings revealed the capability of the proposed scheme to reduce the classifiers’ classification time without much influence on the classification accuracy of the classifier.

With the rapid development of IoT technology in various fields such as smart cities and industry 4.0, the need for wireless sensor network-based systems has increased, one of which is the concept of using a vehicle as a mobile sensor network or known as VaaMSN. Many developers use the IoT platform as a cloud computing service in developing the VaaMSN system. However, not all IoT platform service providers provide monitoring features on every device and provide information such as device location, purpose, condition. Therefore, this research aims to develop an IoT Platform that can receive data and provide information on each device, making it easier to process data and control devices. Therefore, this research aims to develop an IoT platform called the SEMAR-IoT-Platform that able to received data and provide information on each device for easier data processing and control devices. The SEMAR-IoT-Platform integrates Big Data, Data Analytics, Machine Learning, using the principles of Extract, Transfer, and Load (ETL) for data processing and provides communication services using HTTP-POST, MQTT, and NATS. The test results show that the system has been successfully implemented to complement a simple IoT system with an average delay time of HTTP, NATS, and MQTT communications of less than 150ms for the data storage process, and for the data visualization process has an average delay time of less than 300ms.

The purpose of the article is to investigate and consider the general trends, problems and prospects of using cloud computing and technologies in data processing, in particular optimization of their storage. The research methodology consists of semantic analysis methods of the basic concepts of the considered subject area (cloud computing and cloud technologies). The article discusses approaches to storing data (including multimedia) in modern systems that support cloud computing and technologies. The scientific novelty of the research is the analysis of cloud computing use and cloud technologies for data storage optimizing. The conclusions of the research conducted in the article are as follows: the paper considered various aspects related to cloud computing and cloud technologies, including various advantages and disadvantages; cloud computing and cloud technologies can offer numerous benefits to various stakeholders; cloud computing and cloud technologies have some problems; it is noteworthy that by supporting the cloud computing and cloud technologies deployment and maintenance models, problems can be either minimized or eliminated since the dynamic characteristics of cloud computing and cloud technologies can always support users.

Hate speech is one of the most challenging problem internet is facing today. This systematic literature review examine hate speech detection problem and will be used to do an experimental approach on detecting hate speech and abusive language. This work also provide an overview of previous research, including methods, algorithms, and main features used. We use two research questions in this literature review which will be the foundation of the next experimental research. Correctly classifying a piece of text as an actual hate speech requires a lot of correctly labelled data. Most common challenges are different languages, out of vocabulary words, long range dependencies and many more.

Future communication networks may encounter various issues in order to facilitate heavy heterogeneous data traffic and large number of users, therefore more advanced multiple access (MA) schemes is required to meet the changing requirements. Recently, a promising physical-layer MA technique has been suggested for multi-antenna broadcast channels, namely Rate Splitting Multiple Access (RSMA). This new scheme has the ability to partially decode the interference and partially treat the remaining interference as noise which makes it to cope with wide range of user deployments and network loads. On the other hand, interleave division multiple access (IDMA) has already been recognized as a potential code domain NOMA (non-orthogonal multiple access) scheme, suitable for 5G and beyond communication network. Hence, in this paper, a new approach of multiple access scheme is proposed to get the grip on new challenges in future communication (6G). The proposed framework consists the joint processing of RSMA and IDMA (code domain NOMA), in which the transmitter involves an IDMA as encoder and allows rate splitting to split the message in two parts i.e. common part and private part, before the actual transmission. The mathematical modeling of proposed system is elaborated in the paper and for simulation purpose the downlink communication scenario has been considered where users faced diverse channel conditions. The weighted sum rate (WSR) performance is evaluated for the proposed scheme which validate the quality of service (QoS) of the joint RS-IDMA system.

Numerous research and related applications of fuzzy clustering are still interesting and important. In this paper, Fuzzy C-Means (FCM) and Laying Chicken Algorithm (LCA) were modified to improve local optimum of Fuzzy Clustering presented by using UCI dataset. In this study, the proposed FCMLCA performance was also compared to baseline technique based on CSO methods. The simulation results indicate that the FCMLCA method have better performance than the compared methods.

In telecommunication channel modelling the wavelength is small compared to the physical features of interest, therefore deterministic ray tracing techniques provide solutions that are more efficient, faster and still within time constraints than current numerical full-wave techniques. Solving fundamental Maxwell's equations is at the core of computational electrodynamics and best suited for modelling electrical field interactions with physical objects where characteristic dimensions of a computing domain is on the order of a few wavelengths in size. However, extreme communication speeds, wireless access points closer to the user and smaller pico and femto cells will require increased accuracy in predicting and planning wireless signals, testing the accuracy limits of the ray tracing methods. The increased computing capabilities and the demand for better characterization of communication channels that span smaller geographical areas make numerical full-wave techniques attractive alternative even for larger problems. The paper surveys ways of overcoming excessive time requirements of numerical full-wave techniques while providing acceptable channel modelling accuracy for the smallest radio cells and possibly wider. We identify several research paths that could lead to improved channel modelling, including numerical algorithm adaptations for large-scale problems, alternative finite-difference approaches, such as meshless methods, and dedicated parallel hardware, possibly as a realization of a dataflow machine.

Automated segmentation of histopathological images is a challenging task to detect cancerous cells in breast tissue. Recent reviews state high accuracy to segment image, but depends on user input, say window area size, time steps, level set, magnification factor and so on. To extract the region of interest effectively, the subject expert performs post-processing operations several times on the segmentation results with different input values for different parameters say, area opening, fill holes and selects most appropriate enhanced image required for further analysis. The authors proposed an automated segmentation technique followed by self-driven post-processing operations to detect cancerous cells effectively. The post-processing method itself determines the value of different parameters for different operations based on segmented results obtained. The proposed technique has the following features: (i) technique is context sensitive; (ii) no prior setting of time step, weighted area coefficient parameters is required; (iii) magnification independent; (iv) post-processing operations are self-driven which enhance segmentation results adaptively. The experimental results are compared with four state-of-the-art techniques: fuzzy C-means, spatial fuzzy C-means, spatial neutrosophic distance regularised level set and convolutional neural network-based PangNet. Experimental results obtained on two publicly available data sets show that the proposed technique outperforms effectively.

Fuzzy set and neutrosophic set are two efficient tools to handle the uncertainties and vagueness of any real-world problems. Neutrosophic set is more useful than fuzzy set (intuitionistic fuzzy sets) to manage the uncertainties of a real-life problem. This study introduces some new concepts of single-valued neutrosophic graph (SVNG). The authors have discussed the definition of regular SVNG, complete SVNG and strong SVNG. The shortest path problem is a well-known combinatorial optimisation problem in the field of graph theory due to its various applications. Uncertainty is present in almost every application of shortest path problem which makes it very hard to decide the edge weight properly. The main objective behind the work in this study is to determine an algorithmic technique for shortest path problem which will be very easy and efficient for use in real-life scenarios. In this study, the authors consider neutrosophic number to describe the edge weights of a neutrosophic graph for neutrosophic shortest path problem. An algorithm is introduced to solve this problem. The uncertainties are incorporated in Bellman–Ford algorithm for shortest path problem using neutrosophic number as arc length. They use one numerical example to illustrate the effectiveness of the proposed algorithm.

One example of an artificial intelligence ethical dilemma is the autonomous vehicle situation presented by Massachusetts Institute of Technology researchers in the Moral Machine Experiment. To solve such dilemmas, the MIT researchers used a classic statistical method known as the hierarchical Bayesian (HB) model. This paper builds upon previous work for modeling moral decision making, applies a deep learning method to learn human ethics in this context, and compares it to the HB approach. These methods were tested to predict moral decisions of simulated populations of Moral Machine participants. Overall, test results indicate that deep neural networks can be effective in learning the group morality of a population through observation, and outperform the Bayesian model in the cases of model mismatches.

The mystery of aesthetics attracts scientists from various research fields. The topic of aesthetics, in combination with other disciplines such as neuroscience and computer science, has brought out the burgeoning fields of neuroaesthetics and computational aesthetics within less than two decades. Despite profound findings are carried out by experimental approaches in neuroaesthetics and by machine learning algorithms in computational neuroaesthetics, these two fields cannot be easily combined to benefit from each other and findings from each field are isolated. Computational neuroaesthetics, which inherits computational approaches from computational aesthetics and experimental approaches from neuroaesthetics, seems to be promising to bridge the gap between neuroaesthetics and computational aesthetics. Here, we review theoretical models and neuroimaging findings about brain activity in neuroaesthetics. Then machine learning algorithms and computational models in computational aesthetics are enumerated. Finally, we introduce studies in computational neuroaesthetics which combine computational models with neuroimaging data to analyze brain connectivity during aesthetic appreciation or give a prediction on aesthetic preference. This paper outlines the rich potential for computational neuroaesthetics to take advantages from both neuroaesthetics and computational aesthetics. We conclude by discussing some of the challenges and potential prospects in computational neuroaesthetics, and highlight issues for future consideration.

Various machine-learning classification techniques have been employed previously to classify brain states in healthy and disease populations using functional magnetic resonance imaging (fMRI). These methods generally use supervised classifiers that are sensitive to outliers and require labeling of training data to generate a predictive model. Density-based clustering, which overcomes these issues, is a popular unsupervised learning approach whose utility for high-dimensional neuroimaging data has not been previously evaluated. Its advantages include insensitivity to outliers and ability to work with unlabeled data. Unlike the popular k-means clustering, the number of clusters need not be specified. In this study, we compare the performance of two popular density-based clustering methods, DBSCAN and OPTICS, in accurately identifying individuals with three stages of cognitive impairment, including Alzheimer’s disease. We used static and dynamic functional connectivity features for clustering, which captures the strength and temporal variation of brain connectivity respectively. To assess the robustness of clustering to noise/outliers, we propose a novel method called recursive-clustering using additive-noise (R-CLAN). Results demonstrated that both clustering algorithms were effective, although OPTICS with dynamic connectivity features outperformed in terms of cluster purity (95.46%) and robustness to noise/outliers. This study demonstrates that density-based clustering can accurately and robustly identify diagnostic classes in an unsupervised way using brain connectivity.

Alzheimer’s disease (AD) is the most common type of dementia. The exact cause and treatment of the disease are still unknown. Different neuroimaging modalities, such as magnetic resonance imaging (MRI), positron emission tomography, and single-photon emission computed tomography, have played a significant role in the study of AD. However, the effective diagnosis of AD, as well as mild cognitive impairment (MCI), has recently drawn large attention. Various technological advancements, such as robots, global positioning system technology, sensors, and machine learning (ML) algorithms, have helped improve the diagnostic process of AD. This study aimed to determine the influence of implementing different ML classifiers in MRI and analyze the use of support vector machines with various multimodal scans for classifying patients with AD/MCI and healthy controls. Conclusions have been drawn in terms of employing different classifier techniques and presenting the optimal multimodal paradigm for the classification of AD.

In this paper, we estimate Doppler frequency of a maneuvering target being tracked by passive radar using two types of particle filter, the first is “Maximum Likelihood Particle Filter” (MLPF) and the second is “Minimum Variance Particle filter” (MVPF). By simulating the passive radar system that has the bistatic geometry “Digital Video Broadcasting-Terrestrial (DVB-T) transmitter / receiver” with these two types, we can estimate the Doppler frequency of the maneuvering target and compare the simulation results for deciding which type gives better performance.

Cellular networks are constantly evolving to ensure a better Quality of Service (QoS) and quality of coverage ever more important. The radio cellular systems are based on frequency allocation. In this context, frequency allocation principle consists in choosing an optimal frequency plan to meet traffic demand constraints and communication quality while minimizing the radio interferences. This paper proposes an optimal frequency allocation approach based on genetic algorithms to minimize co-channel and adjacent channel interference. The validation of this new approach is confirmed by the results of the work we have done in the GSM network. In fact, we used the file obtained by the OMC-R, which defines the adjacent cells of each cell and the frequencies allocated to the considered area. The results obtained clearly show the effectiveness and robustness of the approach used.

ZigBee is widely used in wireless network in Internet of Things (IoT) applications to remotely sensing and automation due to its unique characteristics compared to other wireless networks. According to ZigBee classification of IEEE 802.15.4 standard, the network consists of four layers. The ZigBee topology is represented in second layer. Furthermore, the ZigBee topology consists of three topologies, star, tree and mesh. Also there are many transmission bands allowed in physical layer, such as 2.4 GHz, 915 MHz, 868 MHz. The aim of this paper is to evaluate the effect of ZigBee topologies on End to End delay and throughput for different transmission bands. Riverbed Modeler is used to simulate multiple ZigBee proposed scenarios and collect the results. The results of the study recommend which topology should be used at each transmission band to provide lowest End to End delay or obtain maximum throughput, which is case sensitive in some IoT applications that required for example minimum delay time or sending high amount of data.

Several technological solutions now available in the market offer the possibility of increasing the independent life of people who by age or pathologies otherwise need assistance. In particular, internet-connected wearable solutions are of considerable interest, as they allow continuous monitoring of the user. However, their use poses different challenges, from the real usability of a device that must still be worn to the performance achievable in terms of radio connectivity and battery life. The acceptability of a technology solution, by a user who would still benefit from its use, is in fact often conditioned by practical problems that impact the person’s normal lifestyle. The technological choices adopted in fact strongly determine the success of the proposed solution, as they may imply limitations both to the person who uses it and to the achievable performance. In this document, targeting the case of a fall detection sensor based on a pair of sensorized shoes, the effectiveness of a real implementation of an Internet of Things technology is examined. It is shown how alarming events, generated in a metropolitan context, are effectively sent to a supervision system through Low Power Wide Area Network technology without the need for a portable gateway. The experimental results demonstrate the effectiveness of the chosen technology, which allows the user to take advantage of the support of a wearable sensor without being forced to substantially change his lifestyle.

This paper presents a detailed review of the applications of augmented reality (AR) in three important fields where AR use is currently increasing. The objective of this study is to highlight how AR improves and enhances the user experience in entertainment, medicine, and retail. The authors briefly introduce the topic of AR and discuss its differences from virtual reality. They also explain the software and hardware technologies required for implementing an AR system and the different types of displays required for enhancing the user experience. The growth of AR in markets is also briefly discussed. In the three sections of the paper, the applications of AR are discussed. The use of AR in multiplayer gaming, computer games, broadcasting, and multimedia videos, as an aspect of entertainment and gaming is highlighted. AR in medicine involves the use of AR in medical healing, medical training, medical teaching, surgery, and post-medical treatment. AR in retail was discussed in terms of its uses in advertisement, marketing, fashion retail, and online shopping. The authors concluded the paper by detailing the future use of AR and its advantages and disadvantages in the current scenario.

Interpretation of brain activity responses using motor imagery (MI) paradigms is vital for medical diagnosis and monitoring. Assessed by machine learning techniques, identification of imagined actions is hindered by substantial intra- and inter-subject variability. Here, we develop an architecture of Convolutional Neural Networks (CNN) with an enhanced interpretation of the spatial brain neural patterns that mainly contribute to the classification of MI tasks. Two methods of 2D-feature extraction from EEG data are contrasted: Power Spectral Density and Continuous Wavelet Transform. For preserving the spatial interpretation of extracting EEG patterns, we project the multi-channel data using a topographic interpolation. Besides, we include a spatial dropping algorithm to remove the learned weights that reflect the localities not engaged with the elicited brain response. We evaluate two labeled scenarios of MI tasks: bi-class and three-class. Obtained results in an MI database show that the thresholding strategy combined with Continuous Wavelet Transform improves the accuracy and enhances the interpretability of CNN architecture, showing that the highest contribution clusters over the sensorimotor cortex with a differentiated behavior of rhythms \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\mu $$\end{document}μ and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\beta $$\end{document}β.

The world is moving towards modernization with the help of smart devices used in smart cities to make the whole lot intelligent and smart. These smart devices are extensively used in smart city and are capable of doing everything which one can do with the normal desktop computer. These smart devices like Smartphone have computational limitations are not able to store a large data to be used and collected in a smart city. In this paper, we propose a novel security architecture which first uses the Elliptic Curve Diffie Helman Key Exchange Algorithms to exchange the keys between the two low power devices (Smartphone). The keys are used to encrypt the large data (images and videos etc). The data is encrypted using the private keys of a device and then send to the cloud for safe storage. The data can be only accessed by the communicating device with the same key. The proposed security architecture enables these smart devices to store the huge data collected from the smart city to store on the cloud. If another device requests the same set of data, the keys can be shared secretly and the communicating device can be allowed to download the data directly from the cloud. This architecture relieves the Smartphone from the storage limitation and also enables it to communicate with faster speed and securely.

Fog Computing is a new concept made by Cisco to provide same functionalities of Cloud Computing but near to Things to enhance performance such as reduce delay and response time. Packet loss may occur on single Fog server over a huge number of messages from Things because of several factors like limited bandwidth and capacity of queues in server. In this paper, Internet of Things based Fog-to-Cloud architecture is proposed to solve the problem of packet loss on Fog server using Load Balancing and virtualization. The architecture consists of 5 layers, namely: Things, gateway, Fog, Cloud, and application. Fog layer is virtualized to specified number of Fog servers using Graphical Network Simulator-3 and VirtualBox on local physical server. Server Load Balancing router is configured to distribute the huge traffic in Weighted Round Robin technique using Message Queue Telemetry Transport protocol. Then, maximum message from Fog layer are selected and sent to Cloud layer and the rest of messages are deleted within 1 hour using our proposed Data-in-Motion technique for storage, processing, and monitoring of messages. Thus, improving the performance of the Fog layer for storage and processing of messages, as well as reducing the packet loss to half and increasing throughput to 4 times than using single Fog server.

Steganography develops tools and methods for hiding the fact of message transmission. The first traces of steganographic methods are lost in ancient times. For example, there is a known method of hiding a written message: the slave's head was shaved, a message was written on the scalp, and after the hair grew back, the slave was sent to the addressee. From detective works, various methods of secret writing between the lines of ordinary text are well known: from milk to complex chemical reagents with subsequent processing. Digital steganography is based on hiding or embedding additional information in digital objects while causing some distortion of these objects. In this case, text, images, audio, video, network packets, and so on can be used as objects or containers. To embed a secret message, steganographic methods rely on redundant container information or properties that the human perception system cannot distinguish. Recently, there has been a lot of research in the field of hiding information in a text container, since many organizations widely use text documents. Based on this, here the MS Word document is considered as a medium of information. MS Word documents have different parameters, and by changing these parameters or properties, you can achieve data embedding. In the same article, we present steganography using invisible Unicode characters of the Space type, but with a different encoding.

The unmanned warehouse dispatching system of the ‘goods to people’ model uses a structure mainly based on a handling robot, which saves considerable manpower and improves the efficiency of the warehouse picking operation. However, the optimal performance of the scheduling system algorithm has high requirements. This study uses a deep Q-network (DQN) algorithm in a deep reinforcement learning algorithm, which combines the Q-learning algorithm, an empirical playback mechanism, and the volume-based technology of productive neural networks to generate target Q-values to solve the problem of multi-robot path planning. The aim of the Q-learning algorithm in deep reinforcement learning is to address two shortcomings of the robot path-planning problem: slow convergence and excessive randomness. Preceding the start of the algorithmic process, prior knowledge and prior rules are used to improve the DQN algorithm. Simulation results show that the improved DQN algorithm converges faster than the classic deep reinforcement learning algorithm and can more quickly learn the solutions to path-planning problems. This improves the efficiency of multi-robot path planning.

Materials discovery is often compared to the challenge of finding a needle in a haystack. While much work has focused on accurately predicting the properties of candidate materials with machine learning (ML), which amounts to evaluating whether a given candidate is a piece of straw or a needle, less attention has been paid to a critical question: Are we searching in the right haystack? We refer to the haystack as the design space for a particular materials discovery problem (i.e. the set of possible candidate materials to synthesize), and thus frame this question as one of design space selection. In this paper, we introduce two metrics, the Predicted Fraction of Improved Candidates (PFIC), and the Cumulative Maximum Likelihood of Improvement (CMLI), which we demonstrate can identify discovery-rich and discovery-poor design spaces, respectively. Using CMLI and PFIC together to identify optimal design spaces can significantly accelerate ML-driven materials discovery., 13 pages, 10 figures, 2 tables

Fluid dynamics simulation is often repeated under varying conditions. This leads to a generation of large amounts of results, which are difficult to compare. To compare results under different conditions, it is effective to overlap the streamlines generated from each condition in a single three-dimensional space. Streamline is a curved line, which represents a wind flow. This paper presents a technique to automatically select and visualize important streamlines that are suitable for the comparison of the simulation results. Additionally, we present an implementation to observe the flow fields in virtual reality spaces.

Integrating artificial intelligence (AI) and computer science together with current approaches in material synthesis and optimization will act as an effective approach for speeding up the discovery of high-performance photoactive materials in organic solar cells (OSCs). Yet, like model selection in statistics, the choice of appropriate machine learning (ML) algorithms plays a vital role in the process of new material discovery in databases. In this study, we constructed five common algorithms, and introduced 565 donor/acceptor (D/A) combinations as training data sets to evaluate the practicalities of these ML algorithms and their application potential when guiding material design and D/A pairs screening. Thus, the best predictive capabilities are provided by using the random forest (RF) and boosted regression trees (BRT) approaches beyond other ML algorithms in the data set. Furthermore, >32 million D/A pairs were screened and calculated by RF and BRT models, respectively. Among them, six photovoltaic D/A pairs are selected and synthesized to compare their predicted and experimental power conversion efficiencies. The outcome of ML and experiment verification demonstrates that the RF approach can be effectively applied to high-throughput virtual screening for opening new perspectives to design of materials and D/A pairs, thereby accelerating the development of OSCs.

Evolutionary algorithms (EAs) and swarm algorithms (SAs) have shown their usefulness in solving combinatorial and NP-hard optimization problems in various research fields. However, in the field of computer vision, related surveys have not been updated during the last decade. In this study, inspired by the recent development of deep neural networks in computer vision, which embed large-scale optimization problems, we first describe a literature survey conducted to compensate for the lack of relevant research in this area. Specifically, applications related to the genetic algorithm and differential evolution from EAs, as well as particle swarm optimization and ant colony optimization from SAs and their variants, are mainly considered in this survey.

An imbalanced dataset is commonly found in at least one class, which are typically exceeded by the other ones. A machine learning algorithm (classifier) trained with an imbalanced dataset predicts the majority class (frequently occurring) more than the other minority classes (rarely occurring). Training with an imbalanced dataset poses challenges for classifiers; however, applying suitable techniques for reducing class imbalance issues can enhance classifiers’ performance. In this study, we consider an imbalanced dataset from an educational context. Initially, we examine all shortcomings regarding the classification of an imbalanced dataset. Then, we apply data-level algorithms for class balancing and compare the performance of classifiers. The performance of the classifiers is measured using the underlying information in their confusion matrices, such as accuracy, precision, recall, and F measure. The results show that classification with an imbalanced dataset may produce high accuracy but low precision and recall for the minority class. The analysis confirms that undersampling and oversampling are effective for balancing datasets, but the latter dominates.

Dengan berkembangnya teknologi informasi menyebabkan efektivitas dalam bekerja meningkat. Dengan adanya website CES (Ciputra Enterprise System) membuat pekerjaan atau tugas kantor dapat dikerjakan di mana saja dan kapan saja. Website yang baik adalah yang bisa meningkatkan keefektifan ketika digunakan dan membuat penggunaan website lebih mudah. Website CES (Ciputra Enterprise System) merupakan sistem baru yang dibuat oleh Ciputra. Namun karena website ini masih baru sehingga masih banyak kekurangan yang dimiliki dari segi kemudahan yang dirasakan pengguna dan efisiensi melalui pengalaman pengguna. Oleh karena itu, penelitian ini bertujuan untuk menganalisis usability dari website CES (Ciputra Enterprise System) untuk meningkatkan usability sistem dengan menggunakan metode USE Questionnaire. Metode ini terdiri dari 4 parameter yaitu Usefulness, Ease of Use, Ease of Learning dan User Satisfaction. Hasil dari penelitian menunjukkan bahwa variabel dari USE Questionnaire yaitu Usefulness, Ease of Use, Ease of Learning berpengaruh signifikan terhadap User Satisfaction secara simultan. Secara parsial, variabel Usefulness, Ease of Use, Ease of Learning berpengaruh signifikan terhadap User Satisfaction. Faktor dominan yang mempengaruhi User Satisfaction (Kepuasan User) dalam menggunakan website CES (Ciputra Enterprise System) adalah Ease of Learning (Kemudahan belajar) dengan nilai koefisien 0,390.

In-hand manipulation is a fundamental ability for multi-fingered robotic hands that interact with their environments. Owing to the high dimensionality of robotic hands and intermittent contact dynamics, effectively programming a robotic hand for in-hand manipulations is still a challenging problem. To address this challenge, this work employs deep reinforcement learning (DRL) algorithm to learn in-hand manipulations for multi-fingered robotic hands. A reward-shaping method is proposed to assist the learning of in-hand manipulation. The synergy of robotic hand postures is analysed to build a low-dimensional hand posture space. Two additional rewards are designed based on both the analysis of hand synergies and its learning history. The two additional rewards cooperating with an extrinsic reward are used to assist the in-hand manipulation learning. Three value functions are trained jointly with respect to their reward functions. Then they cooperate to optimise a control policy for in-hand manipulation. The reward shaping not only improves the exploration efficiency of the DRL algorithm but also provides a way to incorporate domain knowledge. The performance of the proposed learning method is evaluated with object rotation tasks. Experimental results demonstrated that the proposed learning method enables multi-fingered robotic hands to learn in-hand manipulation effectively.

Manually labelling point cloud scenes for use as training data in machine learning applications is a time- and labour-intensive task. In this paper, we aim to reduce the effort associated with learning semantic segmentation tasks by introducing a semi-supervised method that operates on scenes with only a small number of labelled points. For this task, we advocate the use of pseudo-labelling in combination with PointNet, a neural network architecture for point cloud classification and segmentation. We also introduce a method for incorporating information derived from spatial relationships to aid in the pseudo-labelling process. This approach has practical advantages over current methods by working directly on point clouds and not being reliant on predefined features. Moreover, we demonstrate competitive performance on scenes from three publicly available datasets and provide studies on parameter sensitivity.

Diffs, a concept known from source code version control systems such as git, is interesting for geospatial, event-based workflows. We investigate how the native mathematical structure of vector geometries can be utilized in order to create a diffing algorithm tailored to geospatial vector data. Diffing algorithms are a well-researched area which dates to the 1970ies; however, we find that geospatial diffing operations tends to be carried out using generic algorithms combined with a pre- and post-processing step. We created GeomDiff, an algorithm and storage format tailored to geospatial vector data. The creation time, apply/undo time, and patch size of GeomDiff was compared to three other generic algorithms by running an online experiment using 2.5 million real-world geometry pairs from OpenStreetMap. We found that the GeomDiff algorithm performs better than or on-par with the alternatives on point-geometries, and complex geometries with a small (

This study intends to present an innovative study for ranking the alternatives in multiple-criteria decision analysis (MCDA) problems under the interval-valued neutrosophic soft set (IVNSS) environment. In this study, to illustrate the notion of objective and subjective weight, the prospect decision theory (PDT) performs an imperative role in determining decision-making problems. PDT predicts human behaviour in terms of gains and losses and considers the expected utility relation to a reference point rather than complete outcomes. In the analysis of merged criteria weight and the prospect decision-making matrix, the authors get a new dimension level for ranking the array of alternatives. This manuscript provides an improved score function (SF) to convert the interval-valued membership grades of truth, indeterminacy and falsity into a mathematical and computational value. The advantage of this method is that it merges the objective and subjective weight during the proposed method. They propose an algorithm based on the SF to determine MCDA problems with IVNSSs. They illustrate a case study and provide various comparative analyses to show its significance over existing studies.