• 期刊
• OpenAccess

Building an Adaptive Recommendation Model Based on Fuzzy MP Neuron and Weighted Similarity Indicator

Recommender system is one of the most common data filtering techniques used. It helps to discover hidden patterns of information from a wide range of omnipresent products and services. When dataset drifts from scarcity to abundance, the most common methods such as collaborative filtering suffer from information sparsity complication, over-specification, and elevated computational complexity. We have created a hybrid model in this respect that considers between precision and computation time to produce the most appropriate products for customers in real time. We made use of imputation technique, fuzzy logic using novel similarity technique and McCulloch-Pitts (MP) Neuron to cope up with aforementioned complications. The experimental evaluation on MovieLens dataset and comparison with numerous state of art personalization models shows that the proposed model yields high efficiency and effectiveness. We tested the resultant classification accuracy of our proposed model using precision and recall.

• 期刊
• OpenAccess

Cross-Context Semantic Document Exchange Through A Novel Tabular Document Representation Approach

Cross-context semantic document exchange is the process of representing, editing, and transmitting a semantic document in one context, and then receiving and interpreting it in another. It is an important research topic in semantic web, e-commerce and artificial intelligence. The current research methods of semantic document exchange generally include standardization, ontology modeling and collaboration templates, each of which has its own limitations. Based on Tabdoc, any semantic document used for interaction has the same syntax, conceptual meaning, and semantic relationship, regardless of context differences. The new method has been implemented in the SDF Tabdoc based on the XML format. By applying SDF Tabdoc, the Tabdoc editor has been designed and implemented to edit any context-independent document templates and document instances. Finally, the effectiveness of the proposed method is proved by case studies and experiments, indicating that any semantic document can be consistently understood between the parties.

• 期刊
• OpenAccess

On Some New Statistical Correlation Measures for T-spherical Fuzzy Sets and Applications in Soft Computing

Correlation and correlation coefficients are the most utilized statistical tools and measures in the field of engineering, intelligence sciences, data analysis, decision making, biological sciences, etc: In the present communication, we have proposed two new measures of correlation coefficients and measures of weighted correlation coefficients of two T-spherical fuzzy sets based on the newly defined information energy measure under the perception of the four parameters of impreciseness - degree of membership, indeterminacy (neutral), non-membership and the refusal. Further, by implementing the principle of maximum correlation coefficient over the proposed correlation coefficients, the methodologies for solving the problems of pattern recognition and medical diagnosis have been provided with the help of an example for each. A comparative analysis in contrast with the existing methodologies has been presented with comparative remarks and additional advantages.

• 期刊
• OpenAccess

Abnormal Crowd Behaviour Detection Using Parallel Deep Neural Networks

The security of people during public events is one of the significant concerns of authorities. The authorities have to monitor the entire crowd continuously, and they must be capable of preventing all the abnormal activities in the crowd. They are responsible for avoiding these kinds of situations. To prevent such abnormal behaviours, first, they need to detect the abnormal crowd behaviour from the high-density crowd under observation. Detecting abnormal crowd behaviour from the crowd video has been one of the most critical research areas in the intelligent video surveillance system field over the past decade. Numerous strategies for crowd abnormality detection with the assistance of computer vision algorithms and machine learning methods have been proposed in recent years. Many of those traditional approaches are using hand-crafted features like optical flow, HoG, SIFT, and SURF. Even though most of these methods were able to produce a considerably good performance, these methods will take a lot of computational time to extract features, and that enhances the whole computational time. Especially the high-level features like SIFT and SURF are computationally complex, and this will affect the real-time performance of the system. In this paper, we propose a novel deep learning strategy for abnormal crowd behaviour detection. Rather than utilising hand-crafted features, deep neural networks naturally learn feature representations of the crowd video and which will help the system to detect abnormal behaviours. The learned feature representations help the system to differentiate normal and abnormal crowd behaviours. This method uses convolution neural networks, which have been utilised as an integral tool for feature learning in computer vision algorithms to extract the features from the videos. Instead of using traditional single-stream convolution neural networks, we have used pre-trained two-stream convolution neural networks to detect the crowd abnormality, which can consider both spatial and temporal information in the video. Our method is tested with available standard datasets and compared with state-of-the-art methods.

• 期刊
• OpenAccess

An Explainable Diagnostic Method for Autism Spectrum Disorder Using Neural Network

Autism spectrum disorder (ASD), also known as autism, is a mental illness caused by disorders of the nervous system. Autism is mainly characterized by developmental disorders, accompanied by abnormalities in social skills, communication skills, interests, and behavioral patterns. Autism cannot be completely cured by existing medical means, and its symptoms can only be relieved through acquired intervention. The best intervention period for autistic patients is before the age of six. But relying on existing methods, most patients with autism have missed the best intervention period when they are diagnosed. In order to allow the subject to be diagnosed with autism in a timely manner, we proposed a method that uses a deep neural network to analyze the subject's magnetic resonance imaging (MRI) and evaluate the performance for early screening of ASD. Our primary analysis of patients with functional magnetic resonance imaging (fMRI) also compared with structural magnetic resonance imaging (sMRI). Experiments have shown that fMRI is more sensitive to autism than sMRI. In addition, we explain the classification results of fMRI.

• 期刊
• OpenAccess

Outpatient Text Classification System Using LSTM

Outpatient text classification is an important problem in medical natural language processing. Existing research has conventionally focused on rule-based or knowledge-source-based feature engineering, but only a few studies have utilized the effective feature learning capabilities of deep learning methods. A long short-term memory (LSTM) model for the outpatient text classification system was proposed in this research. The system has the ability to classify outpatient categories according to textual content on website Taiwan E Hospital. The experimental results showed that our system has very well in the task. The success of the LSTM model applications in the outpatient system provide users to inquire about their health status as references.

• 期刊
• OpenAccess

Neutrosophic Soft Set and Clinical Application

Neutrosophic soft set is one of the generalizations of classical set theory with parameters. We have introduced weighted similarity measure using the normalized orthogonal distance between two single valued neutrosophic soft sets and their characteristics. Further, a decision-making framework is proposed through an algorithm for multi attribute decision making neutrosophic soft scenario. We also apply the proposed weighted similarity measure to the clinical application; identify the best type of radiotherapy treatment for tumor of moving organs such as lungs or chest walls by evaluating certain medical parameters and computation of mathematical ranking model, which are then compared with other existing similarity measures to illustrate the feasibility of the same.

• 期刊
• OpenAccess

A Space-Time Interactive Visualization Approach for Managing Remote Sensing Data

With the rapid development of earth observation technology, the remote sensing data produced gradually shows characteristics of multi-source and heterogeneous, and data volumes are also exploding. Designing approaches and tools to manage the remote sensing data brings a unique set of research and engineering challenges, specifically with regards to data condition query and interpretation. First, we extract the metadata of the data to prepare for unified data retrieval. Considering the real-time, intuitive and interactive in data management, visual images are generated and used to represent the data itself. By considering the spatial characteristics of remote sensing data, we then propose a method based on mouse real-time plotting for the setting of spatial attribute condition, and the condition can be dynamically modified based on the feedback of the query result. After getting the query result, line frame and wall projection are used to display the data spatial distribution. For the intuitiveness and accuracy of data interpretation, we present an approach based on space-time cube technology to assess in a single view where remote sensing data with different parameters is unobstructed displayed in a layered manner. The approach consists of several parts, including a stack view, two cube boxes and a data probe. As an essential component, an information balloon is used to show data parameters and provide a convenient interface for data manipulation. Finally, we demonstrate the effectiveness and the usefulness of the proposed approach using a real-world case and three tests.

• 期刊
• OpenAccess

Energy Efficient Sustainable City Monitoring Using IoT Enabled Wireless Sensor Networks and Data Analytics

As the population in the city is increasing rapidly over the years, due to the scarcity resources and unhealthy ecosystem the demand for sustainable city increases. Sustainable city enables all its citizens to meet their own needs with minimal natural resources and to live a good quality of life, without degrading the existing natural resources or the lives of other people now or in the future. Once after building sustainable city with green building, energy efficient and eco-friendly ecosystem it is important to monitor the same to keep the city sustainable. An Energy efficient Wireless sensor network with internet connectivity improves the regular monitoring, the frequent data received from various monitoring sensors are considered as informative database for future prediction, these huge information can be used for alerting critical situations through data analytics. The integration of various technology yield performance degradation due the energy usage and computational overhead, which can be improved through application of an optimization technique like genetic algorithm.

• 期刊
• OpenAccess

Seed Optimization Framework on Draughts

Seed optimization has been successfully tested on many games such as Go, Domineering, Breakthrough, among others. Fixed seeds can outperform random seeds by selecting locally optimal seeds as different playing policies. In this article seed optimization has been tested for the Draughts program Scan. We provide a framework which can optimize a draughts program for competition. It does not affect the original program structure, so it improves the strength with no modifying algorithm and no penalty when executing. With the new Best Promise Seed framework, the win rate can be improved by replacing the random seeds with some pretested locally optimal seeds. The optimized program won the championship in the Computer Olympiad in 2015 and 2016. It shows that self learning methodology improves the strength of Scan against other competing programs. In addition, better locally optimal seed(s) may be discovered with a longer learning time, so further strength improvement is possible. All current draughts programs and other different game programs might gain benefit from this framework.