International Journal of Modeling, Simulation, and Scientific Computing

Artificial Intelligence (AI) systems have evolved with digital learning developments to provide thriving soft groups with digital opportunities in response to feedback. When it comes to learning environments, educators’ training feedback is often used as a response recourse. Through the use of final evaluations, students receive feedback that improves their education abilities. To improve academic achievement and explore knowledge in the learning process, this section provides an AI-assisted personalized feedback system (AI-PFS). An individualized feedback system is implemented to learn more about the student’s lack of academic experience interactivity and different collaboration behaviors. According to their benchmark, PFS aims to establish a personalized and reliable feedback process for each class based on their collaborative process and learn analytics modules. It has been proposed to use multi-objective implementations to evaluate students regarding the learning results and teaching methods. With different series of questions sessions for students, AI-PFS has been designed, and the findings showed that it greatly enhances the performance rate of 95.32% with personalized and reasonable predictive.

Recommendation systems are shrewd applications for knowledge mining that profoundly handle the problem of data overload. Various literature explores different philosophies to create ideas and recommends different strategies according to the needs of customers. Most of the work in the suggested structure space focuses on extending the accuracy of the recommendation by using a few possible methods where the principle purpose remains to improve the accuracy of suggestions while avoiding other plan objectives, such as the particular situation of a client. By using appropriate customer rating data, the biggest test for a suggested system is to generate substantial proposals. A setting is an enormous concept that can think of numerous points of view: for example, the community of friends of a client, time, mindset, environment, organization, type of day, classification of an item, description of the object, place, and language. The rating behavior of customers typically varies in different environments. We have proposed a new review-based contextual recommender (RBCR) system application from this line of analysis, in particular a novel recommender system, which is an adaptable, quick, and accurate piece planning framework that perceives the significance of setting and fuses the logical data using piece stunt while making expectations. We have contrasted our suggested calculation with pre- and post-sifting methods as they have been the most common methodologies in writing to illuminate the issue of setting conscious suggestion. Our studies show that considering the logical data, the display of a system will increase and provide better, appropriate and important results on various evaluation measurements.

To estimate the parameters of the model of option pricing based on the model of rough fractional stochastic volatility (RFSV), we have carried out the empirical analysis during our study on the pricing of SSE 50ETF options in China. First, we have estimated the parameters of option pricing model by adopting the Monte Carlo simulation. Subsequently, we have empirically examined the pricing performance of the RFSV model by adopting the SSE 50ETF option price from January 2019 to December 2020. Our research findings indicate that by leveraging the RFSV model, we are able to attain a more accurate and stable level of option pricing than the conventional Black–Scholes (B-S) model on constant volatility. The errors of option pricing incurred by the B-S model proved to be larger and exhibited higher volatility, revealing the significant impact imposed by stochastic volatility on option pricing.

The analysis of the information retrieval system focuses on the notion of appropriate and irrelevant documents. The performance predictor, including accuracy and reminder, is used to establish how well the device satisfies consumer requirements. The effectiveness of the indexing and retrieval is calculated by contrasting a typical collection of queries and documents with the efficacy, functionality, and systemic approach. Important evaluations are used to measure functionality, performance (precession and retraction), compilation, and interface assessments. Document and query indexing, query assessment, and system assessment are key issues in information retrieval. This paper uses the human–computer interaction based on the intelligent information retrieval method (HCI-IRM). The proposed method concentrates on customer satisfaction, the main success assessment metric. It identifies the collection of related records at a given time within the collection. An information retrieval system’s main objective is to obtain the information. It is either the actual information or the documents containing the information substitutes which completely or partly correspond to the customer’s query evaluation. The extraction and recruitment of knowledge-based data from a database are usually related to the retrieval of information. The retrieval and precise technology are used to assess the efficiency of the data recovery system framework. As a result, HCI-IRM enhances the response time, and the relevance of the outcomes is key to customer satisfaction. Comparison of Yahoo and Google search engines focused on accuracy and reminder technology.

Presently, the design process of the AI remote server can enable the user to evaluate whether an authorized user can gain emotional responses when they establish an emotional product interface in the approach of the interaction with the device. Therefore, it is necessary for user experience and the ability to address the user’s emotional expectations. This paper proposes an artificial intelligence-based user face recognition response system (AI-UFRRS) to monitor users’ emotions in real life continually and provide new insights into their emotional responses and transitions. The user face recognition response system design is analyzed based on device intelligence. Eventually, the response system is improved and strategy based on an intelligent device. The proposed AI-UFRRS utilizes reinforcement learning technologies to maintain emotional processing in substantial information relating to the user’s identity. This paper offers AI remote strategies to reduce identification information and maximize information on emotions formed by reinforcement learning. The results suggest that the system provided can perform a convolute transformation to maintain user recognition accuracy and reduce face identity recognition. Thus, the experimental results of AI-UFRRS show the improved accuracy ratio of 95.6%, the recognition rate of 93.4%, emotion ratio of 95.5%, high response system ratio of 96.3%, and to increase user identification ratio of 91.8% and reduced false acceptance rate of 19.2%, the false rejection rate of 19.5% compared to other methods.

In big data, the frequent item set mining is an important framework for many applications. Several techniques were used to mine the frequent item sets, but for the collapsed and complex data, it is difficult. Hence, the current research work aimed to model a novel Frequent Pattern Growth-Hybrid Ant Colony and African Buffalo Model (FPG-HACABM) is developed to overcome this issue and to reduce the execution time. Moreover, the Fitness function of HACABM is utilized to calculate the support count of each item and to improve the classification accuracy. Thus the proposed models classify the frequently utilized items accurately and arranged those items in descending order. This helps to run the big data transactional application effectively without any delay. Finally, the key metrics are validated with the existing models and better results are attained by achieving a high accuracy rate of 99.82% and less execution time of 0.0018 ms.

Radioactive waste sorting often faces an unstructured and locally radioactive working environment. At present, remote operation sorting has problems such as low sorting efficiency, greater difficulty in operation, longer training periods for personnel, and poor autonomous control capabilities. Based on the premise of improving the adaptability and autonomous operation ability of robots in an unstructured environment, this paper uses the dual deep Q learning algorithm to optimize the classic deep Q learning algorithm to improve training speed and improve sorting efficiency and stability. Secondly, the sorting algorithm model of deep reinforcement learning is used to determine the optimal behavior in this state. Set up multiple sets of simulations and physical experiments to verify the sorting method. The results show that the robotic arm can autonomously complete sorting tasks under complex conditions and can significantly improve work efficiency when pushing and grasping collaborative operations and will preferentially grasp objects with high radioactivity in the radioactive area. The algorithm has migration ability and good generalization.

Based on the distributed decision tree algorithm, this paper first proposes a method of vertically partitioning datasets and synchronously updating the hash table to establish an information-based mass data retrieval method in a heterogeneous distributed environment, as well as using interval segmentation and interval filtering technologies for improved algorithm of distributed decision tree. The distributed decision tree algorithm uses the attribute histogram data structure to merge the category list into each attribute list, reducing the amount of data that needs to reside in the memory. Second, we adopt the strategy of vertically dividing the dataset and synchronously updating the hash table, select the hash table entries that can be used to update according to the minimum Gini value, modify the corresponding entries and use the hash table to record and control each sub-site. In the case of node splitting, it has a high accuracy rate. In addition, for classification problems that meet monotonic constraints in a distributed environment, this paper will extend the idea of building a monotonic decision tree in a distributed environment, supplementing the distributed decision tree algorithm, adding a modification rule and modifying the generated nonmonotonic decision tree to monotonicity. In order to solve the high load problem of the privacy-protected data stream classification mining algorithm under a single node, a Storm platform for the parallel algorithm PPFDT_P based on the distributed decision tree algorithm is designed and implemented. At the same time, considering that the word vector model improves the deep representation of features and solves the problem of feature high-dimensional sparseness, and the iterative decision tree algorithm GBDT model is more suitable for non-high-dimensional dense features, the iterative decision tree algorithm will be integrated into the word vector model (GBDT) in the data retrieval application, using the distributed representation of words, namely word vectors, to classify short messages on the GBDT model. Experimental results show that the distributed decision tree algorithm has high efficiency, good speed-up and good scalability, so that there is no need to increase the number of datasets at each sub-site at any time. Only a small number of data items are inserted. By splitting some leaf nodes, a small amount is added by branching to achieve a monotonic decision tree. The proposed system achieves a massive data ratio of 54.1% while compared with other networks of massive data ratio.

In the production process of the communication circulator, defects such as solder joints, missing pieces, and wool may appear due to the current processing technology. The current communication circulator detection relies on manual experience for identification and classification. This method seriously affects production efficiency; and visual fatigue, physical condition, and working environment will affect the accuracy of manual detection; therefore, the production line of communication circulators urgently needs to be online automatic detection system. This paper will design a set of the stable automatic detection system. The main contributions are: (1) In the sample collection stage, by enhancing the limited samples, it proposes the morphological transformation of the limited defects themselves to realize the diversification of the number of samples and the morphology; (2) By reusing a sample, a cascaded Faster-RCNN and You Only Look Once (YOLO) target detection method is proposed, and the Faster-RCNN and YOLO models are trained, respectively; the model training stage uses migration learning and optimization of hyperparameters to obtain higher accuracy high training model; (3) In the inference stage, by cascading the two excellent target detection algorithms, Faster-RCNN and YOLO, this paper proposes a new decision conflict resolution algorithm, and finally obtains a higher comprehensive detection accuracy; (4) In the deployment phase, in view of the increased performance and computing power overhead caused by algorithm cascading, this paper proposes a solution for graphics processing unit shard sharing. Experimental results explain: The communication circulator surface defect detection system based on the deep learning method can effectively locate and identify various defects. The final recognition accuracy rate is about 98.5%, and the recognition speed can reach about 0.45, which is compared with traditional detection methods, which greatly improves the accuracy and speed of recognition.

The forgery involved in region duplication is a common type of video tampering, where the traditional techniques used to detect video tampering are ineffective and inefficient for the forged videos under complex backgrounds. To overcome this issue, a novel video forgery detection model is introduced in this research paper. Initially, the input video sequences are collected from Surrey University Library for Forensic Analysis (SULFA) and Sondos datasets. Further, spatiotemporal averaging method is carried out on the collected video sequences to obtain background information with a pale of moving objects for an effective video forgery detection. Next, feature extraction is performed using the GoogLeNet model for extracting the feature vectors. Then, the Unsupervised Feature Selection with Multi-Subspace Randomization and Collaboration (UFS-MSRC) approach is used to choose the discriminative feature vectors that superiorly reduce the training time and improve the detection accuracy. Finally, long short-term memory (LSTM) network is applied for forgery detection in the different video sequences. The experimental evaluation illustrated that the UFS-MSRC with LSTM model attained 98.13% and 97.38% of accuracy on SULFA and Sondos datasets, where the obtained results are better when compared to the existing models in video forgery detection.

Video and object segmentation are considered significant research topics in image-video processing. The techniques and mathematical models involved in detecting and segmenting objects employ several modules of different high-level approaches developed for video analysis, object extraction, classification, and recognition. Moving object detection is important in various applications like video surveillance, moving object tracking. This paper proposes video segmentation of moving scene using fast convolutional neural network with VGG-16 net architecture which improves the accuracy. This developed method based on CNN sparsely represents foreground, background, and segmentation mask, which is used in reconstructing the original images. The feed-forward network-trained videos are applied for object detection in a single image with co-segmentation approach where videos or image collections are required as the input. The segmentation is performed through comparative analysis of real-time DAVIS dataset. The results of the experiment show the efficiency of this proposed method which is tested and compared with the existing techniques such as convolution neural network, $k$-nearest neighbors, and artificial neural network by the parameters, namely accuracy, precision, recall, and F1-Score. The proposed technique has been improved in terms of accuracy by 97.8%, precision by 94%, recall by 87.9%, and F1-Score by 83.8%.

Unsupervised data clustering investigation is a standout among the most valuable tools and is an informative task in data mining that looks to characterize similar articles’ gatherings. One of the eminent algorithms for the clustering field is K-means clustering. Scholars recommended enhancing the nature of K-means, and optimization algorithms were hybridized. In this study, a heuristic calculation, deer hunting optimization algorithm (DHOA), was adjusted for K-means data clustering by altering the fundamental parameters of DHOA calculation, which are propelled from the characteristic enlivened calculations. During this work, a new human-based descriptive DHOA has been developed following a human deer hunting strategy. In order to attack the fawn, hunters update their positions based on the movement of the leader and backward movement while also considering the angle of the deer. In this work, the DHOA was hybridized with K-means clustering and the performance of the proposed approach is tested against UCI repository data with different algorithms.

A knowledge graph is a visual method that can display the information contained in the knowledge points, core structure, and comprehensive knowledge structure technology. In recent years, with the innovation of science and technology, the business field became keen on knowledge graphs and the graphical display method. However, the application of knowledge graphs in the business field is mainly limited to search engines, question, and answer systems because of the technical difficulties of knowledge extraction and knowledge graph drawing of unstructured text, especially the knowledge extraction of amorphous culture. It can provide knowledgeable service to users by analyzing the knowledge entity contained in encyclopedia knowledge or knowledge base. This paper will focus on the critical link of knowledge extraction of the knowledge graph, adopt a depth learning algorithm to solve this urgent problem and combine with the application of knowledge graph in substation fault to analyze the construction process of substation fault knowledge map based on AI.

Artificial Intelligence (AI), the emergence and development of social system model, are gradually affecting the development of system engineering technology and methods. AI mainly focuses on acquiring new knowledge or skills by looking for rules from observation data (samples); social system model mainly focuses on the ability of description and model demonstration from the current situation to the future. The combination of AI and social system model can promote the research of social problems related to system engineering. The synergetic growth of various sectors of the economic system depends on the initial level of output of each sector and its internal relations. This paper focuses on the internal relationship of the two-sector economic model and explores the impact of distribution relations on the synergetic growth of sectors. The study found that the necessary conditions for synergetic growth is a function of the profit rate, the difference between the labor income (or capital income) of the two sectors, and the marginal consumption propensity. This function has a more complex and nonlinear characteristic. By demonstrating the effects of changes in some parameters, we find that an increase in the profit rate requires that the income distribution structure of the two sectors tend to be the same; an increase in the marginal consumption propensity can allow for greater differences in the income distribution structure of the two sectors.

MobileNet Convolutional Neural Network is a predominately used Deep Learning (DL) model in various computer/mobile vision applications. However, it is clear from the literature that the researches using MobileNet Convolution Neural Network (CNN) reported with only minimal level of classification accuracy in the case of plant nutrient deficiency classification. As a result, on the benchmark rice plant dataset, a modified MobileNet CNN architecture is proposed in this study for nutrient deficiency classification. Insertion of additional new layers, a system to handle dropout and selecting appropriate optimizer are among the alterations included in this proposed method to increase the model performance accuracy. Subsequently, the performance of the modified MobileNet CNN model is evaluated using well-defined performance metrics such as accuracy, ${\mathrm{\text{F}}}_1$-score, precision, and recall. In addition, the performance of the proposed modified MobileNet CNN model is compared with the state-of-the-art conventional DL models such as VGG16, VGG19, ResNet152, ResNet152V2, ResNet101, ResNet101V2, ResNet50, ResNet50V2, InceptionV3, InceptionResNetV2, MobileNet, and MobileNetV2. Empirical evaluation of the proposed model with the cutting-edge DL models shows that the modified MobileNet CNN model outperforms the aforementioned DL models. Finally, we have developed a web and mobile application for automatically classifying nutrient deficiency in rice plants that the farmers could easily use.

Designing efficacious semantics for the dynamic interaction and searches has proven to be concretely challenging because of the dynamically of the semantic searches, method of browsing and visualization interfaces for high volume information. This has a direct impact on enhancing the capabilities of the web. To surmount the challenges of providing meaning to high volume unstructured datasets, Natural language processing techniques and implements have been proven to be propitious, however, the reactivity of these techniques should be studied and predicated on the objective of providing meaning to the unstructured data. This paper demonstrates the working of five NLP techniques namely, bag-of-words, TF-IDF, NER, LSA, and LDA. The experiment provides the kindred attribute accomplishment or the identification of the meaning of this unstructured data varies from one technique to another. However, NLP techniques can be efficient as they provide insights into the data and make it human-readable. This will in turn avail in building better human–machine intractable browsing and applications.

Lung cancer has been identified as the world’s leading cause of death. Precise detection and testing of lung nodules at an early stage are essential for the diagnosis of lung cancer, therefore, it is one of the best ways to prevent deaths of lung cancer. An automated nodule detection system provides a second opinion to radiologists during early diagnosis. Much existing research uses Deep Convolutional Networks (DCNNs) for lung nodules classification. However, DCNNs normally need careful tuning of hyperparameters to reveal their excellent performance. Although, with the increasing size of state-of-the-art convolutional neural networks, the evaluation cost of the traditional optimization algorithms has become deplorable in most cases. Also, Lung CT scans data have a data imbalance problem inherently. So, to solve those problems, this paper proposes IISH-2D DCNN for lung nodule classification. The proposed methodology consists of pre-processing and classification phases. In the pre-processing phase, the 3D-CT scan slice is converted into a 2D-slice, and then the nodule boundary is extracted by calculation of ROI. After that, the extracted boundary is given as input to the IISH-2D DCNN that classifies the lung nodules. The performance of the proposed methodology is compared to the existing works based on accuracy, sensitivity, and specificity metrics. Thus, the proposed model outperforms existing lung nodule classification methodologies with higher accuracy, sensitivity, and specificity that are 99.8%, 97%, and 99%, respectively. Also, the proposed methodology has fewer errors than the state-of-the-art methods. Hence, the suggested method attains better performance in lung nodule classification and proves to be more effective.

Stock market forecast improves the economic development of the country. People invest in the stock market at high risk without performing the appropriate analysis. So, it is essential to identify the stock market value prediction through existing stock market datasets. This analysis can be made through machine learning algorithms. Machine Learning is defined as the process through which only investors will gain the profit when investing the money into the stock market. Many prediction models have been contributed by the existing investigators to estimate the stock value. The contribution by Hybrid Heuristic algorithm is that it provides a coherent future prediction of stock market prices and investors will also get the insights about how to sell and buy a specific stock to get more profit. Gann square technique is used to forecast a price maneuver dependent on value, pattern analysis and period which is also used for intraday online trading without prior knowledge.