Journal of Kufa for Mathematics and Computer

Among different techniques, algorithms and applications of Data Mining, predicting the class
label of unlabeled objects(undefined class label) is a crucial term in the field. The most common
approaches in this area is the use of classification technique (DT, Bayes, SVM, KNN and others)
that represent what is known as supervised learning. However, in many cases no target class
labels and the boundaries are available to perform the prediction, so the new approach
Clustering-classification technique is used.
The work in this paper presents a survey of the most common researches conducted in this field
and discuss their experiments, the algorithms they used, the types of data they utilized, the data
sizes used, and the results they discovered.
According to the results, applying the clustering techniques before classification improved
classification accuracy and reduced experiment execution time. The Cluster Classifier was
proven to be a suitable approach to summarize data by some of the researchers. It achieves a
summarization rate of over 50%, which represents a considerable reduction in the size of the test
datasets .
The findings of the researches indicated that, in addition to feature selection and feature
extraction, data preprocessing (handled missing data and effective outlier detection techniques)
enhanced the classifier performance and accuracy while reducing the classification error.

Quantile regression is one of the methods that has taken a wide space in application in the previous two
decades because of the attractive features of these methods to researchers, as it is not affected by outliers values,
meaning that it is considered one of the robust methods, and it gives more details of the effect of explanatory variables on the dependent variable.In this paper, a Bayesian hierarchical model for variable selection and estimation in the
context of binary quantile regression is proposed. Current approaches to variable selection in the context of binary
classification are sensitive to outliers, heterogeneous values, and other anomalies. The proposed method in this study overcomes these problems in an attractive and direct way.

Abstract— Social media influencers have the power to influence others. Identifying influencers in online social
networks is essential for various applications in many domains such as advertisement, community health
campaigns, administrative science and politics. Detecting influencers on online social networks is
achieved in accordance with specific criteria such as the number of subscribers, the number of interactions with
them, the extent of people’s trust in them, etc. the present
study encompasses different measures such as
application, techniques, dataset, factors, and dataset.
Besides, a table summarising and illustrating the main
ideas and approaches is given.

This paper introduces an experimental study of the heart disease datasets characteristics impact on the
performance of classification algorithms in the aim of identifying the best algorithm for each dataset under its
characteristics. The performance of five machine learning algorithms (logistic regression (LR), K-Nearest Neighbor
(KNN), Decision tree (DT), Random Forest (RF), and support vector machine (SVM)), single layer neural network
(ANN), and deep neural network (DNN), has been evaluated using five heart disease datasets under four data complexity
measurement: number of samples (dataset size), number of features (dimension of dataset), Data sparsity measures,
and correlation of features. All datasets have been processed and normalized then the mutual information-based feature
selection method was used to solve the overfitting problem. The results show that in general, the machine learning
especially the Random Forest algorithm achieves high classification accuracy than deep learning network. In other
hand, the high sparsity and less mutual information of dataset has large impact on degradation of the performance of
classification algorithms than other characteristics of data.

a dataset, one of the most common methods used is clustering. Recent years have seen a rise in the number of articles interested in clustering, which may be attributed to the development of various new fields of application. Most clustering method's performance depends on initial values of parameters or the preprocessing of a dataset, so clustering needs to optimize. The improvement of clustering in two directions, either we improve the parameters or improve the input data. Ga is an evolutionary technique well-known in optimization. This survey deals with the methods that use a genetic algorithm to choose parameter values and input data. This study shows many important performance metrics used to get the optimal result in each research.

In this paper, we introduce a new of a BH-algebra and a new of pseudo BH-algebra. We call this covid-19 of pseudo
BH-algebra. Also, we give the concepts of ( BH-algebra, pseudo BH-algebra and covid-19 ). And we study some relationships between them.

Parameter estimation is a fundamental problem in signal processing. Deep learning is a fundamental
method to solve this problem. This paper used five Deep Learning (DL) methods and three datasets including different
singles Single Tone (ST), Linear- Frequency-Modulated (LFM), and Quadratic-Frequency-Modulated (QFM). This
signal is affected by a mixture of Additive White Gaussian (AWG) noise and Additive Symmetric alpha Stable (SαS)
noise. The recurrent neural network includes Gated Recurrent Unit (GRU), Long Short-Term Memory (LSTM), BiDirection Long
Short-Term
Memory
(BiLSTM), and
Convolution
Neural
Network
includes
1D-CNN
and
2D-CNN.
DL

can

significantly reduce estimation complexity, memory cost, and power consumption as compared to the classical
methods based on time-frequency analysis, which are important requirements for many systems, such as some
Internet of Things (IoT) sensor applications. All DL methods are designed with fewer layers to reduce complexity,
especially in hardware. Geometric SNR (GSNR) is used to determine the impulsiveness of mixture noise in a Gaussian
and SαS noise. When compared to a deep learning classifier with few layers to get on high accuracy and complexity
reduces for Instantaneous Frequency (IF) estimation, Linear Chirp Rate (LCR) estimation, and Quadratic Chirp Rate
(QCR) estimation. IF, LCR, and QFM estimation for ST, LFM, and QFM signals. The results show that 2D-CNN is
better than other deep methods for parameter estimation in LFM signals and QFM signals, and the GRU is better for
parameter estimation in ST signals, where the accuracy of the ST dataset in GRU is 58.09. The accuracy of the LFM
and CFM datasets in 2D-CNN is 98.26 and 98.2 respectively.

In recent years, cloud computing has gained significant popularity, with data storage emerging as a
crucial and valuable aspect of this technology. Cloud data storage refers to transferring data to a remote, up-to-date
storage system rather than relying on a local device such as a computer's hard disc. Cloud storage allows users to
reduce their hardware and software requirements while benefiting from enhanced convenience and cost-effectiveness.
However, security concerns surrounding data access and storage in the cloud often render it unreliable and
untrustworthy. Cloud computing delivers dynamically scalable resources over the Internet, offering numerous
advantages in terms of cost and usability.
Nevertheless, cloud computing services must address security challenges when transmitting sensitive information and
critical applications to shared and public cloud environments. As cloud environments expand to meet data processing
and storage needs, users must carefully consider the potential benefits and drawbacks of these services regarding
data security. This study focuses on the primary security concerns in cloud computing systems, examining the
fundamental issues associated with data protection in these environments. We present a secure, privacy-preserving
architecture for inter-cloud data sharing that employs an encryption/decryption algorithm to safeguard data stored in
the cloud against unauthorized access. By providing a robust and reliable solution to address the security challenges
inherent in cloud computing, this research contributes to developing safer and more trustworthy cloud-based systems
for users.

Localization in Wireless Sensor Networks (WSNs) is process to determine the locations of sensors connected
within a network. Several applications need to determine the location of sensor nodes in the network to help increasing network performance. Some applications, such as item tracking, military operations, routing protocols, robotics, and inventory management, are made feasible by sensor network’s ability to accurately locate the sensor nodes. For each
of these applications, different criteria for position estimation. These are speed, accuracy, and reliability. In this paper
we going to classify the localization algorithms with a new perspective based on three criteria (area of fields,
cooperation between nodes and node’s mobility). Following along this paper may give’s an idea to the researchers to
develop efficient algorithms to localize nodes with accuracy adapting to different techniques with respect to the
geographic area and anchor type to be designed.

Different topological indices will always be valuable in a variety of disciplines, including chemistry,
economics, electronics, business studies, social sciences and medicine. According to the dominating set and number,
the Wiener index for neutrosophic graphs (NG) will be studied in considerable detail in this research.
Additionally, the Average Wiener index for an NG has been defined, and several associated theorems have been
introduced.

Since the COVID-19 epidemic's rise in 2020, Cover face recognize achieve advanced significantly in the
range of computer vision. Face cover is important to stop or limit the COVID-19 disease's spread due to the global
outbreak. Face recognize is among of the most commonly used biometric recognition approach, because it can be
utilized for monitoring systems, identity management, security verifying, and a lot of applications. The majority features
of faces were hidden by mask, leaving just a quite some, including eyes plus head-region, that’s utilized for recognize.
This challenge may reduce the recognition percentage because of the limited area to extract features. Due to the
popularity of deep learning to extract and recognize deep features in many research areas especially computer vision,In
this work, a covered face recognize system is introduced. utilizing Convolutional neural network (CNN), one of the most
widely common deep learning algorithms. The final layer in the CNN architecture, the softmax activation function, was
utilized to identify the facial characteristics after they had been extracted using CNN from the masked face's eyes,
forehead, and brow regions. In the Study employ the "Extended Yale B database," which has issues with changes in
placement and lighting. additionally, they covered faces in Dataset with medical masks. In comparison to other
approaches to solving this problem, our strategy showed to be successful and promising with a recognition accuracy
for "Extended Yale B" of 95%.

In this work, all rings under consideration will be Supposedly to be commutative with nonzero identity
and all modules will be nonzero unite. Restrict Nearly Primary Finitely Compactly Packed Modules is a
generalization of Restrict Nearly Primary Compactly Packed Modules that we introduce. We identify the requirements
that result in the modules being Restrict Nearly primary finitely compactly packed. Also; proved are several results
regarding the Restrict Nearly Primary Finitely Compactly Packed Modules. Investigated are also the pre-and postrequisites for a ???? -module ???? to be Restrict Nearly Primary Finitely Compactly Packed. We
denoted to the submodule by s−mod

The aim of this article is to continue study more spectral properties for class-???????? operators. The Search has
been focused on define new properties for closed linear operators. These new properties posed on class –AN, A. The properties are defined and proved for A. So, the research achieve the goal. Also, the connection between the properties are established.

The rice crop is one of the most important food crops that depend on it globally. Therefore, farmers must
preserve the production of this crop from infection with pests and diseases that lead to its destruction through
artificial intelligence and deep learning techniques. A hybrid model combining a Residual Network 50 (ResNet50)
deep convolutional neural network (CNN) and a support vector machine (SVM) developed diagnoses rice diseases.
Farmers or people working in agriculture could use this model to quickly and accurately identify the diseases in their
crops and treat them, increasing crop yield and reducing the need for costly and time-consuming manual inspection.
ResNet50, a deep learning model effective at image classification tasks, was used to extract features from images of
rice plants. SVM was then used to classify the diseases based on these features. The ResNet50 was able to capture
complex patterns in the images, while the SVM was able to use these patterns to make accurate classification
decisions. This hybrid model allowed for high precision in rice disease diagnosis, achieving an accuracy of
approximately 99%.

Biometric forms major and very effective role nowadays in many fields such as health, reliability, devices,
phones, banking, airport security, and others because of its unique characteristics for each person that cannot be
replicated in another person. Therefore, most security systems rely and verify biometric properties. Airport security
systems rely directly on facial recognition, but these systems may be exposed to attacks by the use of morphing faces in
the passport image that allows multiple users to use the same passport. This paper presents a complete system consist of three stage, the first stage generating morphing faces based on edge detection to determine landmark and combine between landmarks to produce morphing. The second stage passing images on to the face recognition system that using Local Binary Pattern to features extraction, the final stage how to detect image bona fide or morph using texture techniques represented by each Local binary pattern and Gray-Level Co-Occurrence Matrix. With the use of the Wasserstein Distance measure, which has not previously been used in this field. The method gave effective results showing the mechanism of reducing morphing attack. The prposed system is achieved an average accuracy of 75% for AMSL dataset compared with our dataset, with an accuracy of 80%. The power of the proposed work is evident through the FRA and RFF evaluation. Which achieved values as low as possible for dataset FAR 0.25%, indicating the error rate in calculating morphed images is actual, and FRR 0.30, meaning the error rate in calculating the actual images is morphed, these ratios are less than one, the higher system's accuracy in detection. The AMSL dataset ( FAR 0.20 , FRR 27%). It turned out that the training of the proposed method optimized for the features extracted for the landmarks area significantly affects finding the difference and discovering the modified images, even in the case of minor modifications as in the AMSL dataset.

palm print recognition is a biometric technology used to identify individuals based on their unique comfort patterns. Identifying patterns in computer vision is a challenging and interesting problem. It is an effective and reliable method for authentication and access control. In recent years, deep learning approaches have been used for handprint recognition with very good results. We suggest in this paper, a Siamese network-based approach for handprint recognition. The proposed approach consists of two convolutional neural networks (CNNs) that share weights and are trained to extract features from images of handprints, which are then compared using a loss of variance function to determine whether the two images belong to the same person or not. Among the 13,982 input images, 20% are used for testing, 80% for training, and then passing each image over one of two matching subnets (CNN) that transmit weights and parameters. So that, the extracted features become clearer and more prominent. This approach has been tested and implemented using the CASIA PalmprintV1 5502 palm print database, the CASIA Multi-Spectral PalmprintV1 7,200 palm print, and the THUPALMLAB database of 1,280 palm print using MATLAB 2022a. For 13,982 palmprint recognitions in the database, the equal error rate was 0.044, and the accuracy was 95.6% (CASIA palmprintV1, THUPALMLAB, and CASIA Multi-Spectral palmprintV1). The performance of the real-time detecting system is stable and fast enough. The Siamese network approach is particularly useful in cases where there are large numbers of classes with a small number of observations for each, and the network does not need to be retrained when adding or updating data. Overall, the results show that the proposed Siamese network can work with different databases and with unbalanced data, and can provide very good results. Future work can explore the application of this approach to other biometric recognition systems, and investigate the potential of combining it with other deep learning techniques to further improve accuracy and performance. Compared to previous research, the proposed Siamese network achieves very high results with the use of a large and compact database and three different databases. The approach provides a reliable and stable real-time detecting system

In this article we present a new class of modules which is named as a principally ????????????????-lifting modules.
This class termed by Principally ????????????????-lifting in this work which defined as, a module ???? is called Principally ????????????????lifting if for every cyclic submodule ???? of ???? with ????????????????(????) ⊆ ????, there is a decomposition ???? = ℵ⨁???? such that ℵ ≤ ???? and ???? ∩ ???? is g-small in ????. Thus, a ring ???? is called Principally ????????????????-lifting if it is a principally ????????????????-lifting as ????module. We determined it is structure. Several characterizations, properties, and instances are described of these modules'.????

In this paper we introduce the notion of e-gH modules which is a proper generalization of Hopfian modules
and defined as, a module ???? is called an e-gH if any surjective ????-endomorphism ???? of ???? has an e-small kernel, a ring
???? is called an e-gH ring if ???? is an e-gH as R-module. We give some characterizations and properties of this modules.