Journal of Computing and Security
Volume:11 Issue: 1, Winte4 and Spring 2023

This paper introduces a variant version of the AO for efficiently solving high-dimensional computationally expensive problems. Traditional optimization techniques struggle with problems characterized by expensive objective functions and a large number of variables. To address this challenge, this paper proposes a SMAO that leverages machine learning techniques to approximate the objective function. SMAO utilizes RBF to build an accurate and efficient surrogate model. By iteratively optimizing the surrogate model, the search process is directed toward the global optimum while significantly reducing the computational cost compared to traditional optimization methods. To evaluate and compare the performance of SMAO with the surrogate model-based versions of the Gazelle Optimization Algorithm GOA, RSA, PDO, and FLA, they are analyzed on a set of benchmark test functions with dimensions varying from 30 to 200. According to the reported results, SMAO has a higher performance compared to others in terms of achieving the nearest solutions to an optimum, early convergence, and accuracy.

In recent years, artificial intelligence has been used to diagnose and classify cancers using different deep learning-based models. Although they have good overall accuracy, they need high computation resources and execution time, and also have low accuracy, precision, and sensitivity. To this end, we try to employ a new model named the "EfficientNetB0" model with appropriate preprocessing to obtain high precision and sensitivity at a relatively low computation time for diagnosing lung cancer. The EfficientNetB0 model consists of 7 blocks, and each block includes one layer of mobile convolution (MB-Conv) and squeeze-excitation (SE) blocks. EfficientNetB0 has a higher accuracy compared to other common deep learning models due to incorporating a compound coefficient approach. The proposed model is evaluated on the histopathology images dataset and the obtained accuracy of the model is 0.9258. Also, its precision and sensitivity are 0.942 and 0.967, respectively, and these show the superiority of this model compared to existing methods.

Influence maximization in social networks has been an important research issue in the recent decade. This issue is identifying the most influential individuals in a social network who can convey influence to the largest number of the network’s members. However, in influence maximization, the costs of all of the nodes to be selected as seeds are considered the same for the companies that do not hold in the real world. Accordingly, influence maximization-cost minimization has gained attention recently. Available studies have applied multi-objective optimization methods which are time-consuming. Applying the existing approaches of influence maximization for other variants of this problem has been considered in some studies about other multi-objective versions of the influence maximization problem. In this study, extending and well-applying run time-efficient methods of influence maximization are considered to influence maximization-cost minimization. Accordingly, two methods are proposed. The first, Local Lowest Degree Rank (LLDR) is a heuristic-based one which by considering the degree of nodes aims to find the cost-affordable influential nodes with minimum influence overlap among them. The second proposed method, Ratio-aware-CELF-based (RCELF) method, is a Cost Effective Lazy Forward (CELF)-based algorithm which extends CELF as a run-time efficient greedy approach for influence maximization by incorporating the cost function of the nodes into consideration. The proposed methods are evaluated by applying two real-world datasets, Facebook and Last.fm. The results establish the outperformance which in comparison with the most effective benchmark method is between 4% to 32% for LLDR and between 34% to 96% for RCELF.

In recent years, emotion recognition as a new method for human-computer interaction has attracted the attention of researchers. Automatic speech emotion recognition has become one of the practical methods to increase engagement in most industries. It is expected that emotion recognition based on audio information can result in better accuracy. The purpose of this article is to present an efficient method for recognizing emotional states from speech signals, based on a new cognitive model. Due to the importance of the topic, this article presents an efficient method for recognizing emotional states from speech signals based on a mixed deep learning and cognitive model called SOAR. To implement each part of this model, two main steps have been introduced. The first step is reading the video and converting it to images and preprocessing it. The next step is to use the combination of convolutional neural network (CNN) and learning automata (LA) to classify and detect the rate of facial emotional recognition. The reason for choosing CNN in our model is that no dimension is removed from the speech signal and considering the temporal information in dynamic speech leads to more efficient and better classification. Also, the training of the CNN network in calculating the backpropagation error is adjusted by LA so that the efficiency of the proposed model is increased and the working memory part of the SOAR model can be implemented. In the proposed model, audio databases available in the field of multimodal emotion recognition eNTERFACE' 05 and SAVEE have been used for various experiments. The recognition accuracy of the presented model in the best case from eNTERFACE' 05 and SAVEE databases is equal to 85.3% and 84.5%, respectively.

Cloud-based machine learning has become an increasingly popular approach for
training and deploying machine learning models, thanks to its scalability, cost-effectiveness, and ease
of access. However, the use of cloud-based machine learning also introduces new security and privacy
challenges, particularly with respect to insider threats. In this proposed research project, we aim to
develop a multi-faceted approach to enhancing security and privacy in cloud-based machine learning.
Our approach will draw on a range of techniques, including fully homomorphic encryption, multi-factor
authentication. The proposed framework conducts a comprehensive evaluation using a variety of
datasets and use cases, and this approach provides higher security and privacy as compared to existing
security and privacy frameworks for cloud-based machine learning. The ultimate goal is to provide
practical and effective solutions for enhancing security and privacy in cloud-based machine learning,
and to contribute to the ongoing efforts to address the challenges of insider threats in this rapidly evolving field.

Text analysis has been one of the issues in recent research to identify users' sentiments. Most studies have identified sentiments' positive and negative polarity in Persian, and limited research has been done on analyzing emotions in Persian sentences by covering the primary emotional states. In this study, first, a dataset of emotional sentences was prepared to label six basic emotional states, JAMFA. This dataset contains 2350 sentences and (31222 words). This paper presents two models, efficient BERT-BiLSTM(EBB) and XLM-R Catboost(XLM-RC), that enhance the performance of the Persian text emotion classification. This study has the advantages of human intelligence methods and statistical approaches to achieve better accuracy in sentence labeling. The evaluation indicates the accuracy of labeling is 92%, and the reliability of the dataset based on the type of emotions is 88%. The results show that the models at best achieved 86\% accuracy in basic emotion classification and an 81% F-score in binary classification.