Search

The powerful yet less recognized field of fractal dynamics is studied by the author for its origin and applications. Some naturally observed fractals are studied and categorized. Simple ways of creating a fractal is discussed along with the step-by-stem images. One new fractal is also developed and calculated for its surface area in this paper. The infinite surface area bounded in visibly finite region is the notable property of a fractal which can be used to explain many mysterious phenomena in nature. The proposed paper is a small effort to categories fractals in two different ways and giving a thought which is only a conceptual transcript which may need theoretical and practical evidences later on. [ABSTRACT FROM AUTHOR]

The reliable operation of power systems is essential to prevent over-voltages, excessive currents, outages, and equipment damage, which can pose risks to users and lead to interruptions in the power supply. To address these challenges, a robust power safety system is required to swiftly identify, categorize, and pinpoint faults, ultimately reducing their impact. This paper presents a comprehensive approach for locating, classifying, and detecting faults within power systems. The primary objectives of this paper are to accurately identify and categorize different fault types occurring at various resistance levels and locations, gain insights into the causes of interruptions, expedite power restoration, and minimize recurrence. Additionally, this research aims to enhance understanding of the components of the protection system, enabling the implementation of preventive measures to reduce the likelihood of service interruptions and equipment damage. To achieve precise problem identification, classification, and fault localization, the proposed approach employs simple neural networks and efficient machine learning models. Leveraging the capabilities of neural networks, this technology yields precise results in power system fault analysis. [ABSTRACT FROM AUTHOR]

In the modern age, a significant number of official documents, including IDs approved by the government and certificates, are predominantly image-based or paper-based. Extracting, entering, and searching for information from these documents manually can be a cumbersome and time-consuming process for organizations. To address this challenge, Automated ID Document Extraction and Classification (AIDEX) is introduced. This model utilizes machine learning algorithms to automatically classify identity documents issued by the government into predetermined categories such as AADHAAR card, Driving License and PAN card. By identifying unique indicators present in these ID proofs, AIDEX effectively detects and displays vital information, eliminating the need for manual intervention and achieving automation. Through appropriate access and permissions, the extracted data can be securely saved and accessed by authenticated users as required. The application of this strategy significantly reduces the time spent on physical labor, leading to resource conservation. Notably, AIDEX stands out by integrating the identification and data extraction from images of Driving Licenses, distinguishing itself as a novel approach. This paper provides a comprehensive overview of the objectives set for AIDEX, the various techniques employed during its development, and the technical specifications of the resulting model. The findings highlight the potential of AIDEX to streamline information retrieval processes, enhance operational efficiency, and promote automation in handling government-issued identity documents. [ABSTRACT FROM AUTHOR]

Intelligent models that tackle classification problems have shown that it can be developed using different algorithms that are available to use. Therefore, adding an additional level of complexity during the construction of such models. This paper focused on implementing Naive Bayes classification algorithm to the Optimized Electronic Nose System (OENS). Furthermore, the current implementation of the classification layer in OENS is by the usage of the SVM classification algorithm. This paper implemented and presented the results of the implementation of Naive Bayes algorithm by using the same dataset used during the development of OENS. In addition, the development of the Naive Bayes model was done by the usage of the WEKA tool. Apart from that, this paper served as an experiment to compare the results the Naive Bayes algorithm provided compared to the SVM algorithm used during the development of OENS. The classification model that was developed using Naive Bayes algorithm had an accuracy of 89.7% which is lower than the current accuracy of OENS classification model which is 98.1%. Therefore, the implementation of SVM for this problem has been deemed to be better than Naive Bayes. [ABSTRACT FROM AUTHOR]

In the present day, deep learning applications are widely used in every field of development. In this paper, such a type of application is discussed in real-time. Mura dataset is the largest abnormal dataset given by the University of Stanford. In this research paper, we will understand the pre-processing of the image dataset. Kera's and TensorFlow libraries are used to build this model. We are classifying images using different ML models where accuracy is compared to get the best model build. This paper gives us a clear understanding of building deep learning models, understanding pre-processing techniques, and working on heavy data. This research's main agenda is to create a model which can classify bone fracture images and also give a small path for further development in the deep learning area. In this project we are using Resnet50, InceptionV3, Mobile Net and DenseNet with optimizer Adam and comparing their accuracies graphically. [ABSTRACT FROM AUTHOR]

We consider genetic and evolution algebras generated by non-ergodic Volterra operator. It is known that a zero-sum game generated by Volterra operator be a RPS game if and only if the operator is a non-ergodic transformation. We will call the genetic (evolution) algebra generated by non-ergodic Volterra operator RPS genetic (respectively RPS evolution) algebra. In this paper, we investigate the problem of isomorphism of two RPS genetic (evolutionary) algebras and establish necessary and sufficient conditions when two such algebras will be isomorphic. [ABSTRACT FROM AUTHOR]

The paper proposes a novel approach for recognizing and classifying gym exercises using human pose detection with Mediapipe and LSTM, followed by matching the accuracy of the exercises using the Hungarian algorithm and cosine similarity. The proposed approach involves the use of a combination of cutting-edge technologies, including computer vision and deep learning, to recognize and classify a range of gym exercises accurately and efficiently. This study presents several advantages of this approach, including the ability to leverage temporal information, robustness to variations in lighting, camera angle, and occlusions, and the ability to handle a wide variety of gym exercises. Experimental results demonstrate that the proposed approach achieves high accuracy and outperforms existing methods for gym exercise recognition and classification. The paper also compares the proposed approach with existing techniques, highlighting its strengths and cost-effectiveness. It concludes by outlining the methodology, including dataset collection and pre-processing, landmark extraction, LSTM model training, exercise classification, and similarity analysis. Finally, the paper discusses potential future research areas in this field. [ABSTRACT FROM AUTHOR]

This paper focuses on fault classification in cascaded H bridge nine-level inverter (CHBMLI). Comparative fault analysis and fault classification are presented. Primitive methods of fault classification like wavelet and principal component analysis were used to classify faults in multi-level inverters. Wavelets were widely used for fault classification as it gives the details of time and frequency information simultaneously. With the advent of neural networks and other soft computing techniques, multilevel inverter fault analysis has become much simpler. Healthy output is compared with faulty signals and this is fed to the neural network. A Neural network trains the signals fed into the network. As a result, the performances of faulty and healthy signals are studied, and results are presented using MATLAB/SIMULINK. [ABSTRACT FROM AUTHOR]

Enhanced power graphs are important graphs that are extensively studied in characterizing finite groups due to their practical and interesting properties. The properties of the graphs are related to the selection of a set of vertices in their definition, which can be emphasized as one of the developments of the graph. Different sets of vertices selected will lead to the different prop- erties of the graphs. Graph invariants and classification of the graph on groups are essential properties of the defined graph. Graph invariants are the graph properties that remain unchanged after certain transformations, which serve as computational approaches in several areas. Additionally, the classification of the graph is used to establish the defined graphs by the characteristics of the group they represent. A general graph presentation of the desired groups can help researchers to find properties more efficiently. Therefore, this paper highlights the general presentation, some graph invariants, and classification of the deep enhanced power graph of the semi-dihedral groups. The deep enhanced power graph of a finite group G is the graph whose vertices are the elements G except the nontrivial central element, and two different vertices are adjacent if they belong to the same proper cyclic subgroup. It is constructed for all finite semi-dihedral groups to discover their patterns, followed by general presentation formation. Some numerical graph invariants of the deep enhanced power graphs of the semi-dihedral groups are obtained using the general presen- tation: the degree of a vertex, clique number, chromatic number, independence number, domination number, girth, and diameter. For all semi-dihedral groups, the deep enhanced power graph is classified as a connected and perfect graph. [ABSTRACT FROM AUTHOR]

Text classification using K Nearest Neighbor, which locates the K closest matches in training data and uses the label of closest matches to forecast the class of the text. Euclidean distance is used to find the closet matchwhich is predict the class level. The technique Bag of Words is used to find numerical values from the document in terms of text representation. An annotated corpus in Odiabased on agriculture domain collected from ILCI, Govt. of India treated as training data and the representation of that text data for finding sentiment using the concept of the feature vector made from the entire document which we have collected. The calculation of the results using KNN with tf-idf is challenging for Odia text as it low resource language. This paper presents details of result analysis with thoroughly done by experiment. Evaluation of the model is focusedon speed of the transferring text data to the proposed model and getting high accuracy rate. The result of the testing data depends on good and bad features of the algorithm and suggests for how to improve feature iteration of the same model in future. [ABSTRACT FROM AUTHOR]

Fingerprint is one of the most famous biometric characteristics. It is employed in many fields such as forensic, security, recognition and classification. This paper focuses on clustering fingerprint images into original and fake. Unsupervised Deep Leaning (UDL) is proposed, it exploits the Self-Organization Maps (SOM) to provide such clustering. It consists of two internal processing parts. The first part is for the feature extraction. The second part is for the unsupervised clustering of the SOM. Fingerprint images from the ATVS-FakeFingerprint DataBase (ATVS-FFpDB) for without cooperation are utilized in our work. Multiple clustering and classification metrics of the Silhouette Value (SV), Calinski Harabasz Index (CHI), Davies-Bouldin Index (DBI) and accuracy are provided. Also, different comparisons with state-of-the-art Deep Learning (DL) architectures are provided. Our UDL approach has achieved a high accuracy result of 92.86% and fingerprint images are successfully clustered into original and fake categories. [ABSTRACT FROM AUTHOR]

Quantum machine learning is a new field of study that aims to apply the advantages of quantum computing, such as superposition, which is one of the intriguing properties that could speed up computation, to machine learning. Like classical machine learning models, quantum machine learning models can learn from specific input data and perform specific prediction tasks. However, quantum machine learning models are currently in the form of quantum circuits with parameterizable gates. In this paper, the implementation of a quantum classification model is investigated using the two most well-known quantum machine learning development tools: Cirq with TensorFlow Quantum (TFQ) and Qiskit with PyTorch. The quantum classification circuit is trained to recognize images of handwritten digits using the MNIST dataset. However, because the quantum computer is still in development and is difficult to access and use, quantum simulators are the best alternative for the time being. The performance of the quantum classification model implemented by the two development tools is then compared in this paper experimentally using quantum simulators on two different hardware configurations. In addition, the experimental setups differ in the number of input qubits used to store the image's classical pixel values. Finally, the wall time and CPU time used to run the training process, as well as classification accuracy, are examined. In terms of accuracy, Qiskit with PyTorch outperforms Cirq with TFQ only in the lower number of qubits but the results are similar in the higher number of qubits. The experiment results in terms of run-time are highly different. From the result, we find that Cirq is noticeably faster than Qiskit in every experiment by at least 90 times. [ABSTRACT FROM AUTHOR]

Fault analysis, detection and protection are fundamental to avert unpredicted results in solar photovoltaic (PV) systems. Regardless of the verity that solar PV systems have no stirring components and typically involve small upholding, they are still topic of matter to different malfunctions or errors along the PV arrays, power habituation components, batteries, wiring, and functional interconnections. This paper is presenting a review analysis of various articles presented by earlier researchers in the field of fault detection, classification, and protection for solar PV system. In solar photovoltaic (PV) systems, fault analysis, detection, and protection are essential to prevent unexpected outcomes. Solar PV systems are subject to various faults or defects throughout the PV arrays, power habituation components, batteries, wiring, and functional linkages despite the fact that they have no moving parts and normally require little maintenance. This work reviews and analyses a several number of earlier academics' papers in the area of fault classification, detection, and protection for solar PV systems. [ABSTRACT FROM AUTHOR]

Within the last twenty years, the world has witnessed a data explosion, and Arabic is the world's fifth most used textual communication language. The main challenge for document classification is a lack of label information; unstructured textual data is available everywhere. It is obvious that human label assignment is time-consuming and labor-intensive. Automated processes alone cannot assign class labels to a large number of texts while simultaneously updating the classification model. However, in scenarios with a few labeled examples and a large number of unlabeled examples, semi-supervised machine learning (SSL) can significantly enhance performance. SSL combines labeled and unlabeled data to build a more accurate classification model. Any supervised algorithm can be quickly adapted to use an SSL approach using a self-learning algorithm. As SSL in this paper, this work converted seven traditional classifiers from supervised behavior into a semi-supervised one; moreover, it contributed the use of the co-training algorithm on Arabic text, again using the seven classifiers, it used two different views rather than dividing one view into two views, each view being used to train the cooperating classifier. To increase the reliability of our work, experiments have been done on three reliable global datasets (SANAD, OSAC, and CNN). The results have been calculated for precision, recall, F-Score, and accuracy, the performance reached more than 97%. [ABSTRACT FROM AUTHOR]

Due to its extraordinarily contagious nature, the Corona virus disease (COVID-19), which initially started in December 2019, has seriously imperiled the world's healthcare systems. One of the most destructive global pandemics in history is due to its great infectiousness and high mortality rate. As a result, it is crucial to implement a variety of techniques to quickly identify the Corona virus. With the enhancement of early diagnostic and disease severity detection technologies the load on the healthcare structure can be lessened and the effectualness of treatment and rehabilitation for patients can be increased. In light of the severity classification of COVID 19 procedures, this survey analyses numerous methodologies. 25 research papers that focused on different techniques will be used in this study, and an overview of the techniques will be given. The analysis will then be highlighted in the work based on the year of publication, performance metrics, software utilized, datasets and success of the study approaches toward various severity levels of COVID 19and the most utilized techniques related to COVID-19 severity classification is deep learning techniques. At last, drawbacks and issues involved in COVID 19 severity classification techniques will be considered concluding declaration and future scope. [ABSTRACT FROM AUTHOR]

Selection of the most significant variables, i.e. the wavenumber, from an infrared (IR) spectrum is always difficult to be achieved. In this preliminary paper, the feasibility of genetic algorithms (GA) in identifying most informative wavenumbers from 150 IR spectra of papers was investigated. The list of selected wavenumbers was then employed in Linear Discriminant Analysis (LDA). GA procedure was repeated 30 times to get different lists of variables. Then the performances of LDA models were estimated via leave-one-out cross-validation. A total of six to eight wavenumbers were identified to be valuable variables in the GA procedures. All the 30 LDA models achieve correct classification rates between 97.3% to 100.0%. Therefore the GA-LDA model could be a suitable tool for differentiating white papers that appeared to be highly similar in their IR fingerprints. [ABSTRACT FROM AUTHOR]

The infrared (IR) spectra of different white copy paper types tend to appear indifferent. Discrimination between white copy papers could lead to the solution of forgery cases. In this preliminary study, three varieties of white paper were purchased from local stationery shops in Kuala Lumpur, Malaysia. The papers were classified according to their manufacturers using Partial Least Squares-Discriminant Analysis (PLS-DA) and sparse PLS-DA (sPLS-DA) models. The error rate for the two models on the training and the test data sets were estimated and compared. Results show that the performance of the two models are comparable. By including the first six latent variables in both models, classification accuracy as high as 100% can be achieved. [ABSTRACT FROM AUTHOR]

Texture classification proposed in this paper develops deep learning (DL) models with Transfer learning (TL) and fine-tuning method on a cloud-based platform. In this paper, the lightweight deep learning architectures are compared and investigated between MobileNetV2 and NASNetMobile based on transfer learning technique. The standard texture database KTH-TIPS2b, and Kylberg Texture database v1.0 are used for classification. These datasets are split up into three different random ratios as training, validation, and testing. For each split ratio, two different Mobile Deep Learning (MDL) architectures are evaluated, whose weights are trainable from ImageNet. The observations infer the best-split ratio based on performance among the pre-trained models. These models are classified using SoftMax classifier. An Adam Optimization algorithm was utilized to enhance the model's performance and to accelerate the learning speed. As the experiments are conducted with gray and color texture databases, the findings demonstrate 97% accuracy with the best-split ratio that greatly reduces memory usage and computation. [ABSTRACT FROM AUTHOR]

In this paper, Convolutional Neural Network Segmentation (CNNS) by Dual-Interface Reign Architecture is proposed to figure out the condition of breast tumor. The existing novel approaches from machine learning may not be voluntarily useful as they require a huge training data bit with segmented pixels. This architecture is built upon a CNN and a dual mode network which is used to segment and can non-linearly maps the input data into a deep latent space respectively. the two paths of dual-interface networks are locality preserving learner and a conditional graph learner, both are used for the extraction of images and exploit the input features. The complemented dual-interface architecture is further used to improve the rate of classification. This proposed method can overwhelm the restriction of some benchmarks for the classification in breast imageries. In addition, to perform extensive error analysis, an analytical model is developed for mammograms and tested for varying pixel widths and input probabilities. The analytical model is validated through simulation. The results of our proposed method demonstrate that it can effectively detect the different kinds of mammograms in database by segmenting them more precisely. When comparing with previous works the precision is improved up to 5% and 8% than the GSA-ANN and ANN respectively. [ABSTRACT FROM AUTHOR]

5G communications use mm-wave communications integrated with massive MIMO systems to meet the high data rates and low delay requirements. For mm-wave propagation, some challenges appear due to the high path loss of the mm-wave. Beam alignment is a considerable condition for wireless communications, which may cause a massive overhead, adversely impacting initial access, handover, and tracking. Selecting the practical beam alignment technique impact good links. In this paper, we offer a review of beam alignment or beam training techniques classifications used to train the Beam in mm-wave communications. It helps to detect the ways for future search in cellular communications in 6G networks. We will be delighted if this short review provides some guidelines for the researchers at mm-wave communications. [ABSTRACT FROM AUTHOR]

A blood smear is used for a diagnostic is important and crucial to detect a group of diseases, therefore the accurate diagnosis of the disease plays a vital role in the planning and evaluation of treatment. The correct classification of blood cells makes the detection of the disease more accurate and reliable. The cell morphology has the main role in detecting and identifying types of blood cells, especially sickle cells because the shape of sickle cells differs greatly from normal cells. This paper reviews the recent methods used in diagnosing sickle cell anemia. Blood cells detection and classification as well as their limitations are also discussed. [ABSTRACT FROM AUTHOR]

In this paper, Modify Lyapunov-Schmidt method (MLSM) is applied to find the key function conforming to the Fredholm functional. Critical points for key function of codimensions 121 has been classified by calculating the geometric description of caustic. The level curves for this function are derived and sketched in the space of parameters. [ABSTRACT FROM AUTHOR]

Because of its simplicity and ease of implementation, the Nave Bayes algorithm is widely employed in text categorization. Multivariate Bernoulli and multinomial models are the most common event models utilized by Nave Bayes for text categorization. A huge number of research use the multinomial model plus Laplace smoothing simply because they believe It outperforms the multivariate model under practically all circumstances. This paper examines Nave Bayes event models and smoothing approaches from a new view in order to throw some light on this generally held belief. To further understand the differences between Nave Bayes events models, their classification performance is examined on datasets in English and Turkish. The findings of our comprehensive trials reveal that the multinomial model does not always perform better. When paired with a proper smoothing strategy, a multivariate Bernoulli model, on the other hand, can perform well under a variety of size criteria for training data. [ABSTRACT FROM AUTHOR]

Generally, a set of frames represents a video structure, clips, or scenes. A segmentation process including breaking down a video sequence into its main components should pre-processed video analysis-based classification methods. Recently, Deep Learning Models-based video analysis and classification approaches have been grown-up and developed to be more concise and convenient for modern technologies such as big data, cloud computing, video surveillance, and video summarization systems. This paper focuses on the knowledge related to deep learning-based methods to achieve object detection and tracking along with video sequences. Our revision presents and discusses various studies of video classification tasks. Further, the fundamental purpose of this research is to look at which of these techniques affected mainly the performance of video classification tasks and the main parameters required to design and implement an efficient video classification system with relative challenges. A comparison study is performed on various types of video classification models to highlight the strong points of each model with a comprehensive analysis of its performance evaluation based on accuracy metrics. [ABSTRACT FROM AUTHOR]

In this paper, our aim is to classify the brain tumour grades innovatively through machine learning using linear regression. Machine learning using linear regression and image pre-processing algorithms are implemented to detect grade classification of tumours. Twenty samples are taken as a sample size with p=0.8 and has been used to improve detection rate in terms of accuracy and sensitivity. The linear regression algorithm has achieved 93% of accuracy and 0.154 standard deviation while image pre-processing has achieved 90% of accuracy, significance <0.01 and also 0.092 standard deviation. Within the limitations of study, linear regression algorithms have higher significance than image processing for the selected data. [ABSTRACT FROM AUTHOR]

One of the most burning issues is to secure the information that is available on the cloud. Now a day's services like storage, platform, virtualization, resource management, and software services that are available on the cloud and connected through the internet, and handling of those devices on a public platform is having high risk. There may be malicious nodes that increase the malicious activities. Preventing such kind activities of cloud services should achieve security goals. The malicious node in general can perform two types of attack one is active and the other is a passive attack due to the open access. Some gen-eral issues faced in cloud computing environments are security management, administrative issues, access controls, and security awareness. Since in cloud environment, a huge amount of data and various resources are available. That can be accessed by mul-tiple users simultaneously. Hence it is required to provide some explicit mechanism to control such access. In this paper, we study various cloud access control management features and outcomes. We are expecting to evaluate comparisons of access control models and classification methodology to give a direction for resolving a problem in the current methodology. This sur-vey highlights challenges and gives direction to researchers so they introduce some novel approaches and mechanisms to prevent the information and other resources from any physical damage and unauthorized access. [ABSTRACT FROM AUTHOR]

To predict the accuracy percentage of Short Message Services (SMS) spam detection using machine learning classifiers. Two ensemble learning algorithms named random forest algorithm and naive bayes are applied to data. The algorithms have been implemented and tested over a dataset which consists of 5574 records. Ensemble learning methods combined several models trained with a given learning algorithm to improve accuracy. After performing the experiment as result shows mean accuracy of 89.75 % by using random forest and compared naive bayes algorithm mean accuracy is 88.05% for SMS spam detection. There is a statistically significant difference in accuracy for two algorithms is p<0.05 by performing independent samples t-tests. This paper is intended to implement Innovative Classification for prediction of SMS spam detection. The comparison results shows that the naive bayes algorithm has appeared to be better performance than Random Forest Algorithm. [ABSTRACT FROM AUTHOR]

Data Mining is one of the greatest critical aspects of automatic disease diagnosis and disease prediction. It includes data mining algorithms and techniques to study medicinal data. Recently, liver disorders have frequently increased and liver illnesses have become one of the greatest deadly diseases in different countries. In this paper, we collected 534 instances of Iraqi liver patients from Baqubah Teaching Hospital. This collected dataset includes several elements such as Age, Total Bilirubin, Direct Bilirubin, Gender, Aspartate Aminotransferase, Total Proteins, and Albumin and Globulin ratio. This paper explores the early prediction of liver disease using various data mining algorithms. The main aim of this paper is to compute the performance of many data mining techniques and compare their performance with the performance of The Indian Liver Patient Dataset. We applied three classification algorithms and parameters like accuracy, precision, recall, and f-measure were investigated to evaluate the performance of these classifiers. [ABSTRACT FROM AUTHOR]

k Nearest Neighbour (kNN) method is a frequently implemented method for improve the pattern classification and solve the problem of imbalanced dataset that has discovered so many classification and clustering applications. In this paper, issues of classification were focused and changes to the nearest neighbouring method were suggested which exploit data from a dataset structure. The issue of class imbalance has recently attracted considerable attention from researchers in data mining strategies. The results of the present experiments using unique client identifier (UCI) repository datasets show that the classifiers generated the work much effective when compared with classic kNN & are much accurate, but not dramatically slow. This produced some fascinating and encouraging outcomes, which inspired more work to develop the kNN process. The purpose of this paper is to suggest versions of kNN which are appropriate for categorization of issues, since it leverage intrinsic of knowledge within data sets, like the composition of the datasets. [ABSTRACT FROM AUTHOR]

Malicious programs have created a major threat in the area of cyber security. Malware detection and classification is a big challenge for the researchers. Now days Machine Learning techniques using Dynamic analysis of a malicious file play an important role for malware detection. Some new type of malware as polymorphic and metamorphic cannot detected easily. Their tactic hide them from anti malware system, such type of malware creates new instance and encrypting the malicious payload as well as changing the code structure at each infection, while retaining the same functionality. To address this we purpose a model for Polymorphic and metamorphic malware detection. This paper addresses detection and classification problem by providing a deeper analysis of API calls, key features and their parameters that enable polymorphism in malware. We named this model as MPDC, This paper also proposed a Feature Engineering approach for the better classification of malware family, this research is based on behavioral (Dynamic) features analysis and API. We used 8 type of malware family for classification. Our model achieved a Detection accuracy rate of 98.74%, and malware family classification accuracy rate of 96%. This research will revolutionize anti-malware industry in creating better protection mechanisms. [ABSTRACT FROM AUTHOR]

Simple linear models, have manifest spectacular execution to form document representations in recent past. In this paper, we put forward the conception of containers by experiments and theoretical affirmations. We perceive that the charter repository and the chronicle matrix indisputably unable to entirely crammed beyond the canister and will no longer have some well-formed and syntactic statistics on exceedingly huge narrative datum. We also advance a well-organized proposition for document delineation, by using clustering algorithms to split up a repository container into various sub-containers and initiate the correlation in the middle of sub-containers. We moreover dispute the possessions of both types of clustering algorithms, DVEM-K means and DVEM-Random, on vast content datasets by tenderness inquiry and topic ranking endeavours. Set side by side to simple even models, the outcomes manifest that our model give-rise to finest document depictions for document-eminent division correspondence tasks. Our perpectives can also be set forth to another models based on neural webbings, such as convolutional semantic networks, repetitive interconnected networks and productive antagonistic networks, in superintend or semi-supervised settings. [ABSTRACT FROM AUTHOR]

The performance assessment of students is the most important work as this is directly related to the future of the students. This research paper focuses on structural study, analysis and identification of non academic characteristics and data mining techniques, specifically classification and clustering. These are required for designing a model which collectively gives performance assessment in the form of academic and non-academic characteristics. These characteristics need to be classified separately on the basis of different subcategories in each type of characteristics and data collected from the questionnaire from the stakeholders of this model students, teachers and parents. Identifying and using classification and clustering algorithms suitable to the structure of this assessment leads to required output of the research. [ABSTRACT FROM AUTHOR]

The ensemble classification technique is a paradigm that combines several classifiers to improve classification performance. One of the most popular ensemble techniques is majority voting. This technique suffers from the fact that it gives a role to the number of classifiers regardless of the accuracy of the classification, and thus completely reflects the behaviour of the majority, whether it is true or false. This paper proposed a classifier combination method based on a majority vote. Whereas the validation and testing results are relied upon after training to give weights to classifiers, the best of those that have a better weight take a greater role in the classification process. The model was tested on three textual datasets with various languages, sizes, and numbers of classes. The proposed method gave better results than the individual classifiers and the traditional majority voting method in all types of data. The proposed method achieved a slight outperformance rate of close to 1%. [ABSTRACT FROM AUTHOR]

Rescue agencies like fire departments, police departments, and health departments need to identify the important and emergency messages to provide better services to the people. In generally those departments will receive a lot of messages which consist of several categories like spam, invalid etc. These departments need a system that can categorize the received messages into categories like important, invalid and spam. This paper proposes a methodology using a naïve Bayes algorithm integrated with a natural language toolkit. The proposed methodology classifies the received messages into alter1, alert2, alert3, invalid and spam messages. Depending upon the category of messages the rescue agencies will take necessary action. The proposed methodology helps the rescue agencies in saving time by taking action against important messages and ignoring unimportant messages. The system is trained through the preprocessed data. This will help the system in categorizing the test data more accurately. The accuracy of the classification is compared with the existing methodologies. The proposed methodology showed 97% accuracy in the classification of the messages into the different categories. [ABSTRACT FROM AUTHOR]

Now days, the requirement for summarized and classified information has been increased because, in this era of the internet, a lot of unstructured data is flowing rapidly. Legal documents are already available in a complex form for the ordinary person, and it increases its complexity when it comes to the internet in various variants. The variation of legal documents includes multiple types of legal cases with longlengths of documents, multiple languages etc. Itmotivates the researchers to work on the automatic classification of legal documents. This paper discusses the detailed work done on legal document classification using deep learning techniques. Also, we compare the results of various models based on slandered evaluation metrics. The results show that the PRTIRCDNL approach outperformed document classification and achieved 96.17% recall, 95.9% F-measure, 95.65% precision and 92.5% accuracy. [ABSTRACT FROM AUTHOR]

In this paper, classification and parameter prediction problems in vibrating cracked nanobeams are considered. The size-dependent nonlocal elasticity theory developed by Eringen is applied. The crack is modelled as massless rotational and longitudinal spring which connects the adjacent segments of the nanobeam. The spring is introduced to take into account the additional strain energy caused by the presence of a crack. The databases of modal data for inverse problems are calculated numerically. The classification and regression problems are studied by k-nearest neighbors, decision trees, and random forest algorithms. The importance of features in patterns is studied as well. [ABSTRACT FROM AUTHOR]

Remote sensing multi-band hyperspectral image classification is a challenging task and is of current interest too. Proper extraction of spectral features of the data results in higher classification accuracy. This paper proposes a hybrid approach to obtain maximum classification accuracy of hyperspectral data. First and foremost, the Fast Fuzzy Cluster Means (FastFCM) algorithm is presented for overcoming the limitation of high computational time of conventional FCM for high-dimensional data. The segmented outcome as spectral information is then combined with the spatial information obtained by the Extended Multiattribute Profiles (EMAPs) to form a composite kernel (filter). In this experiment, the Support Vector Machine (SVM) is used as a classifier to train the data. The classification outcome in terms of three parameters Overall Accuracy (OA), Average Accuracy (AA), and Kappa Index (KI) of the proposed approach (named as (FastFCM-SVM)) has been compared with some current state-of-art methods to investigate its better stability and classification performance. [ABSTRACT FROM AUTHOR]

This paper is about the classification, up to isomorphism, of two-dimensional algebras over any basic field. It provides a full classification of such algebras in terms of their matrices of structure constants which is helpful in solving many other problems related to two-dimensional algebras. [ABSTRACT FROM AUTHOR]

Nowadays social media plays a significant role in all sorts of our activities ranging from analysing the attitude of a person for the job, getting opinions towards buying a product, acting as a forum for exchanging thoughts about the current events of various domains, creating awareness to the public about the natural calamities, educating the public about the fraudulent news spread by the fakers, initiating the young aspirant to protest against any societal issues, etc. Grasping the opinions shared by the experienced people towards a product, film, event, news, or politics like any subject of matter is one among the worth noting applications for a common man. It extends its application to making decisions about our day-to-day activities. The text reviews consist of enormous, sparse, non-uniform distribution of words represented as features. Text mining is the backend process for those applications. It includes techniques such as feature representation, sentiment classification, feature optimization, etc. Analysing the opinions suggested by the experienced people as positive and negative reviews is a challenging process and it is the baseline of our work. This paper contributes to the related processes involved in analysing the sentiments from the text reviews and accurately classifying them based on their polarity. In the proposed work, we particularly focus on feature representation techniques that have a major effect on enhancing the performance of sentiment classification. We explore different feature representation models such as TF-IDF vectorizer, word2vec vectorizer, and glove vectorizer as these word embedding models are interpreting the words and their syntactic and semantic relationships differently from the corpus. Also, we employ machine learning algorithms and a deep convolution neural network to perform comparative studies in classifying the sentiments. The word2vec in combination with Deep Convolution Neural Network provides the accuracy of 85.7%, precision of 84.4%, recall of 87%, and F-measure of 85.7% compared to other models. [ABSTRACT FROM AUTHOR]

This paper is about the classification, up to isomorphism, of two-dimensional algebras over any basic field. It provides a full classification of such algebras in terms of their matrices of structure constants which is helpful in solving many other problems related to two-dimensional algebras. [ABSTRACT FROM AUTHOR]

Software developers in modern, large-scale software engineering should focus on writing code and continually examine monitoring data to justify dynamic behavior in their systems. In recent years, in light of the culture of development operations (DevOps), the monitoring responsibility of the developer has emerged. However, code violations may occur due to lack of experience, short deadlines, and market competition. This study discusses monitoring challenges and suggests improvements to improve the performance and quality of the software product so that developers can take advantage of these improvements in the future. This paper discusses code smell detection techniques, detection tools, classification, and how to refactor the code. In this review, we have presented several different techniques that detect code odor in the literature. These techniques were analyzed by examining the advantages, disadvantages, limitations, and efficiency of these techniques in code odor detection. A comparison was made of some advantages and disadvantages of the techniques used in odor detection and the limitations of the challenge. Finally, a solution was proposed to improve the performance of the odor detection process of the source code and thus improve the quality of the program produced by suggesting PyCos in odor detection for code through mining techniques Data contributing to the development of the Integrated Development Environment (IDE). [ABSTRACT FROM AUTHOR]

Deep learning is an area emerged by the capabilities of machine learning and artificial intelligence. In the field ofdeep learning, various parameters are essential to building an efficient model. These parameters include activation function,loss function, number of layers, number of neurons, and many other essentials. Finding an optimal workable environment bysetting all these variables is really time consuming and challenging due time constraints or projects having sharp deadlines. In this paper, the work is done for the exploration of these parameters that make an environment in which deep learning models work. The experimental results will also be discussed to learn about the best environment that make text classificationoptimal in deep learning. These results would help the professions and researchers to extract a ready to work environment for the projects associated with the text classification. [ABSTRACT FROM AUTHOR]

Discretization is a technique of changing numerical values to categorical values. Discretization plays an important role in classification. Nowadays we are dealing with huge amount of numerical continuous data. Many of the classification algorithms performs best when categorical data is fed to it. So discretization is gaining importance nowadays. This paper compares the performance of discretization on the accuracy of Random Forest classifier. The variables are discretized based on the correlation of variables with the 'outcome' and 'disorder'. The correlation is identified using heatmap. It is found that the discretization of most correlated variables improves the accuracy of Random Forest Classification. But the Random Forest Classification performs best upon the discretization of continuous numerical attributes. For performing the experiment two publicly available datasets are chosen. In both datasets there are continuous attributes. Upon discretizing those continuous attributes, the performance of Random Classifier further increased. But when discretized all the attributes together degraded the performance of Random Classifier. The Discretization method used is Equal Width Binning Discretization [6]. When all the continuous data are discretized the performance is degraded, due to over discretization [1]. The datasets chosen for the experiment are Pima Indian Diabetes Dataset and Liver Disease Prediction Dataset [21]. [ABSTRACT FROM AUTHOR]