Search results for: graph-based feature filtering method

Authors: Yuyan Yi, Jingyi Zheng, Nedret Billor

Abstract:

Independent component analysis (ICA) is one of the most commonly used blind source separation (BSS) techniques for signal pre-processing, such as noise reduction and feature extraction. The main parameter in the ICA method is the number of independent components (IC). Although there have been several methods for the determination of the number of ICs, it has not been given sufficient attentionto this important parameter. In this study, wereview the mostused methods fordetermining the number of ICs and providetheir advantages and disadvantages. Further, wepropose an improved version of column-wise ICAByBlock method for the determination of the number of ICs.To assess the performance of the proposed method, we compare the column-wise ICAbyBlock with several existing methods through different ICA methods by using simulated and real signal data. Results show that the proposed column-wise ICAbyBlock is an effective and stable method for determining the optimal number of components in ICA. This method is simple, and results can be demonstrated intuitively with good visualizations.

Keywords: independent component analysis, optimal number, column-wise, correlation coefficient, cross-validation, ICAByblock

Procedia PDF Downloads 78

Authors: Amir Hajian, Sepehr Damavandinejadmonfared

Abstract:

In this paper the issue of dimensionality reduction is investigated in finger vein recognition systems using kernel Principal Component Analysis (KPCA). One aspect of KPCA is to find the most appropriate kernel function on finger vein recognition as there are several kernel functions which can be used within PCA-based algorithms. In this paper, however, another side of PCA-based algorithms -particularly KPCA- is investigated. The aspect of dimension of feature vector in PCA-based algorithms is of importance especially when it comes to the real-world applications and usage of such algorithms. It means that a fixed dimension of feature vector has to be set to reduce the dimension of the input and output data and extract the features from them. Then a classifier is performed to classify the data and make the final decision. We analyze KPCA (Polynomial, Gaussian, and Laplacian) in details in this paper and investigate the optimal feature extraction dimension in finger vein recognition using KPCA.

Keywords: biometrics, finger vein recognition, principal component analysis (PCA), kernel principal component analysis (KPCA)

Procedia PDF Downloads 340

Authors: Chao Yu, Qing-Guo Wang, Dan Zhang

Abstract:

The outlier detection problem for dynamic systems is formulated as a matrix decomposition problem with low-rank, sparse matrices and further recast as a semidefinite programming (SDP) problem. A fast algorithm is presented to solve the resulting problem while keeping the solution matrix structure and it can greatly reduce the computational cost over the standard interior-point method. The computational burden is further reduced by proper construction of subsets of the raw data without violating low rank property of the involved matrix. The proposed method can make exact detection of outliers in case of no or little noise in output observations. In case of significant noise, a novel approach based on under-sampling with averaging is developed to denoise while retaining the saliency of outliers and so-filtered data enables successful outlier detection with the proposed method while the existing filtering methods fail. Use of recovered “clean” data from the proposed method can give much better parameter estimation compared with that based on the raw data.

Keywords: outlier detection, system identification, matrix decomposition, low-rank matrix, sparsity, semidefinite programming, interior-point methods, denoising

Procedia PDF Downloads 287

Authors: Daleesha M. Viswanathan, Sumam Mary Idicula

Abstract:

Dynamic hand gestures are an intrinsic component in sign language communication. Extracting spatial temporal features of the hand gesture trajectory plays an important role in a dynamic gesture recognition system. Finding a discrete feature descriptor for the motion trajectory based on the orientation feature is the main concern of this paper. Kalman filter algorithm and Hidden Markov Models (HMM) models are incorporated with this recognition system for hand trajectory tracking and for spatial temporal classification, respectively.

Keywords: orientation features, discrete feature vector, HMM., Indian sign language

Procedia PDF Downloads 342

Authors: Yagya Raj Pandeya, Joonwhoan Lee

Abstract:

The dynamic behavior of music and video makes it difficult to evaluate musical instrument playing in a video by computer system. Any television or film video clip with music information are rich sources for analyzing musical instruments using modern machine learning technologies. In this research, we integrate the audio and video information sources using convolutional neural network (CNN) and pass network learned features through recurrent neural network (RNN) to preserve the dynamic behaviors of audio and video. We use different pre-trained CNN for music and video feature extraction and then fine tune each model. The music network use 2D convolutional network and video network use 3D convolution (C3D). Finally, we concatenate each music and video feature by preserving the time varying features. The long short term memory (LSTM) network is used for long-term dynamic feature characterization and then use late fusion with generalized mean. The proposed network performs better performance to recognize the musical instrument using audio-video multimodal neural network.

Keywords: multimodal, 3D convolution, music-video feature extraction, generalized mean

Procedia PDF Downloads 189

Authors: Mohammadreza Hedyehzadeh, Mahdi Yousefi

Abstract:

Colon cancer is one of the most common cancer, which predicted to increase its prevalence due to the bad eating habits of peoples. Nowadays, due to the busyness of people, the use of fast foods is increasing, and therefore, diagnosis of this disease and its treatment are of particular importance. To determine the best treatment approach for each specific colon cancer patients, the oncologist should be known the stage of the tumor. The most common method to determine the tumor stage is TNM staging system. In this system, M indicates the presence of metastasis, N indicates the extent of spread to the lymph nodes, and T indicates the size of the tumor. It is clear that in order to determine all three of these parameters, an imaging method must be used, and the gold standard imaging protocols for this purpose are CT and PET/CT. In CT imaging, due to the use of X-rays, the risk of cancer and the absorbed dose of the patient is high, while in the PET/CT method, there is a lack of access to the device due to its high cost. Therefore, in this study, we aimed to estimate the tumor size and the extent of its spread to the lymph nodes using MR images. More than 1300 MR images collected from the TCIA portal, and in the first step (pre-processing), histogram equalization to improve image qualities and resizing to get the same image size was done. Two expert radiologists, which work more than 21 years on colon cancer cases, segmented the images and extracted the tumor region from the images. The next step is feature extraction from segmented images and then classify the data into three classes: T0N0، T3N1 و T3N2. In this article, the VGG-16 convolutional neural network has been used to perform both of the above-mentioned tasks, i.e., feature extraction and classification. This network has 13 convolution layers for feature extraction and three fully connected layers with the softmax activation function for classification. In order to validate the proposed method, the 10-fold cross validation method used in such a way that the data was randomly divided into three parts: training (70% of data), validation (10% of data) and the rest for testing. It is repeated 10 times, each time, the accuracy, sensitivity and specificity of the model are calculated and the average of ten repetitions is reported as the result. The accuracy, specificity and sensitivity of the proposed method for testing dataset was 89/09%, 95/8% and 96/4%. Compared to previous studies, using a safe imaging technique (MRI) and non-use of predefined hand-crafted imaging features to determine the stage of colon cancer patients are some of the study advantages.

Keywords: colon cancer, VGG-16, magnetic resonance imaging, tumor size, lymph node metastasis

Procedia PDF Downloads 36

Authors: Hee-Jeong Ahn, Kee-Won Kim, Seung-Hoon Kim

Abstract:

The knowledge of the relationship between characters can help readers to understand the overall story or plot of the literary fiction. In this paper, we present a method for extracting the specific relationship between characters from a Korean literary fiction. Generally, methods for extracting relationships between characters in text are statistical or computational methods based on the sentence distance between characters without considering Korean linguistic features. Furthermore, it is difficult to extract the relationship with direction from text, such as one-sided love, because they consider only the weight of relationship, without considering the direction of the relationship. Therefore, in order to identify specific relationships between characters, we propose a statistical method considering linguistic features, such as syntactic patterns and speech verbs in Korean. The result of our method is represented by a weighted directed graph of the relationship between the characters. Furthermore, we expect that proposed method could be applied to the relationship analysis between characters of other content like movie or TV drama.

Keywords: data mining, Korean linguistic feature, literary fiction, relationship extraction

Procedia PDF Downloads 346

Authors: Haojie Wang, Xun Li, Rui Yin

Abstract:

In order to solve the memorization overfitting in the meta-learning MAML algorithm, a method of generating mutually exclusive tasks based on data augmentation is proposed. This method generates a mutex task by corresponding one feature of the data to multiple labels, so that the generated mutex task is inconsistent with the data distribution in the initial dataset. Because generating mutex tasks for all data will produce a large number of invalid data and, in the worst case, lead to exponential growth of computation, this paper also proposes a key data extraction method, that only extracts part of the data to generate the mutex task. The experiments show that the method of generating mutually exclusive tasks can effectively solve the memorization overfitting in the meta-learning MAML algorithm.

Keywords: data augmentation, mutex task generation, meta-learning, text classification.

Procedia PDF Downloads 66

Authors: Amhimmid Mohammed Saffour, Abdelkader Salama

Abstract:

Filtering is a part of image enhancement techniques, it is used to enhance certain details such as edges in the image that are relevant to the application. Additionally, filtering can even be used to eliminate unwanted components of noise. Medical images typically contain salt and pepper noise and Poisson noise. This noise appears to the presence of minute grey scale variations within the image. In this paper, different filters techniques namely (Median, Wiener, Rank order3, Rank order5, and Average) were applied on CT medical images (Brain and chest). We using all these filters to remove salt and pepper noise from these images. This type of noise consists of random pixels being set to black or white. Peak Signal to Noise Ratio (PSNR), Mean Square Error r(MSE) and Histogram were used to evaluated the quality of filtered images. The results, which we have achieved shows that, these filters, are more useful and they prove to be helpful for general medical practitioners to analyze the symptoms of the patients with no difficulty.

Keywords: CT imaging, median filter, adaptive filter and average filter, MATLAB

Procedia PDF Downloads 293

Authors: Haojie Wang, Xun Li, Rui Yin

Abstract:

In order to solve the memorization overfitting in the model-agnostic meta-learning MAML algorithm, a method of generating mutually exclusive tasks based on data augmentation is proposed. This method generates a mutex task by corresponding one feature of the data to multiple labels so that the generated mutex task is inconsistent with the data distribution in the initial dataset. Because generating mutex tasks for all data will produce a large number of invalid data and, in the worst case, lead to an exponential growth of computation, this paper also proposes a key data extraction method that only extract part of the data to generate the mutex task. The experiments show that the method of generating mutually exclusive tasks can effectively solve the memorization overfitting in the meta-learning MAML algorithm.

Keywords: mutex task generation, data augmentation, meta-learning, text classification.

Procedia PDF Downloads 100

Authors: Saad Bakkali

Abstract:

Image processing is a powerful tool for the enhancement of edges in images used in the interpretation of geophysical potential field data. Arial and terrestrial gravimetric surveys were carried out in the region of Tangier-Tetuan. From the observed and measured data of gravity Bouguer gravity anomalies map was prepared. This paper reports the results and interpretations of the transformed maps of Bouguer gravity anomaly of the Tangier-Tetuan area using image processing. Filtering analysis based on classical image process was applied. Operator image process like logarithmic and gamma correction are used. This paper also present the results obtained from this image processing analysis of the enhancement edges of the Bouguer gravity anomaly map of the Tangier-Tetuan zone.

Keywords: bouguer, tangier, filtering, gamma correction, logarithmic enhancement edges

Procedia PDF Downloads 399

Authors: Bikshalu Kalagadda, Srikanth Rangu

Abstract:

The important object of images segmentation is to extract objects with respect to some input features. One of the important methods for image segmentation is Level set method. Generally medical images and synthetic images with low contrast of pixel profile, for such images difficult to locate interested features in images. In conventional level set function, develops irregularity during its process of evaluation of contour of objects, this destroy the stability of evolution process. For this problem a remedy is proposed, a new hybrid algorithm is Clustering Level Set Evolution. Kernel fuzzy particles swarm optimization clustering with the Distance Regularized Level Set (DRLS) and Selective Binary, and Gaussian Filtering Regularized Level Set (SBGFRLS) methods are used. The ability of identifying different regions becomes easy with improved speed. Efficiency of the modified method can be evaluated by comparing with the previous method for similar specifications. Comparison can be carried out by considering medical and synthetic images.

Keywords: segmentation, clustering, level set function, re-initialization, Kernel fuzzy, swarm optimization

Procedia PDF Downloads 328

Authors: Afshin Kadri

Abstract:

Using Renewable Energy Resources (RES) as a type of DG unit is developing in distribution systems. The connection of these generation units to existing AC distribution systems changes the structure and some of the operational aspects of these grids. Most of the RES requires to power electronic-based interfaces for connection to AC systems. These interfaces consist of at least one DC/AC conversion unit. Nowadays, grid-connected inverters must have the required feature to support the grid under sag voltage conditions. There are two curves in these conditions that show the magnitude of the reactive component of current as a function of voltage drop value and the required minimum time value, which must be connected to the grid. This feature is named low voltage ride-through (LVRT). Implementing this feature causes problems in the operation of the inverter that increases the amplitude of high-frequency components of the injected current and working out of maximum power point in the photovoltaic panel connected inverters are some of them. The important phenomenon in these conditions is ripples in the DC bus voltage that affects the operation of the inverter directly and indirectly. The losses of DC bus capacitors which are electrolytic capacitors, cause increasing their temperature and decreasing its lifespan. In addition, if the inverter is connected to the photovoltaic panels directly and has the duty of maximum power point tracking, these ripples cause oscillations around the operating point and decrease the generating energy. Using a bidirectional converter in the DC bus, which works as a buck and boost converter and transfers the ripples to its DC bus, is the traditional method to eliminate these ripples. In spite of eliminating the ripples in the DC bus, this method cannot solve the problem of reliability because it uses an electrolytic capacitor in its DC bus. In this work, a control method is proposed which uses the bidirectional converter as the fourth leg of the inverter and eliminates the DC bus ripples using an injection of unbalanced currents into the grid. Moreover, the proposed method works based on constant power control. In this way, in addition, to supporting the amplitude of grid voltage, it stabilizes its frequency by injecting active power. Also, the proposed method can eliminate the DC bus ripples in deep voltage drops, which cause increasing the amplitude of the reference current more than the nominal current of the inverter. The amplitude of the injected current for the faulty phases in these conditions is kept at the nominal value and its phase, together with the phase and amplitude of the other phases, are adjusted, which at the end, the ripples in the DC bus are eliminated, however, the generated power decreases.

Keywords: renewable energy resources, voltage drop value, DC bus ripples, bidirectional converter

Procedia PDF Downloads 45

Authors: Shahabeddin Sotudian, M. H. Fazel Zarandi, I. B. Turksen

Abstract:

Hepatitis is one of the most common and dangerous diseases that affects humankind, and exposes millions of people to serious health risks every year. Diagnosis of Hepatitis has always been a challenge for physicians. This paper presents an effective method for diagnosis of hepatitis based on interval Type-II fuzzy. This proposed system includes three steps: pre-processing (feature selection), Type-I and Type-II fuzzy classification, and system evaluation. KNN-FD feature selection is used as the preprocessing step in order to exclude irrelevant features and to improve classification performance and efficiency in generating the classification model. In the fuzzy classification step, an “indirect approach” is used for fuzzy system modeling by implementing the exponential compactness and separation index for determining the number of rules in the fuzzy clustering approach. Therefore, we first proposed a Type-I fuzzy system that had an accuracy of approximately 90.9%. In the proposed system, the process of diagnosis faces vagueness and uncertainty in the final decision. Thus, the imprecise knowledge was managed by using interval Type-II fuzzy logic. The results that were obtained show that interval Type-II fuzzy has the ability to diagnose hepatitis with an average accuracy of 93.94%. The classification accuracy obtained is the highest one reached thus far. The aforementioned rate of accuracy demonstrates that the Type-II fuzzy system has a better performance in comparison to Type-I and indicates a higher capability of Type-II fuzzy system for modeling uncertainty.

Keywords: hepatitis disease, medical diagnosis, type-I fuzzy logic, type-II fuzzy logic, feature selection

Procedia PDF Downloads 279

Authors: Hanlin Liu, Hua Li, Yintan AI

Abstract:

Network performance prediction (NPP) is essential for the management and optimization of software-defined networking (SDN) and contributes to improving the quality of service (QoS) in SDN to meet the requirements of users. Although current deep learning-based methods can achieve high effectiveness, they still suffer from some problems, such as difficulty in capturing global information of the network, inefficiency in modeling end-to-end network performance, and inadequate graph feature extraction. To cope with these issues, our proposed Graphormer-based architecture for NPP leverages the powerful graph representation ability of Graphormer to effectively model the graph structure data, and a node-edge transformation algorithm is designed to transfer the feature extraction object from nodes to edges, thereby effectively extracting the end-to-end performance characteristics of the network. Moreover, routing oriented centrality measure coefficient for nodes and edges is proposed respectively to assess their importance and influence within the graph. Based on this coefficient, an enhanced feature extraction method and an advanced centrality encoding strategy are derived to fully extract the structural information of the graph. Experimental results on three public datasets demonstrate that the proposed GraphNPP architecture can achieve state-of-the-art results compared to current NPP methods.

Keywords: software-defined networking, network performance prediction, Graphormer, graph neural network

Procedia PDF Downloads 16

Authors: C. V. Aravinda, H. N. Prakash

Abstract:

In this paper, we tried to convey fusion and state of art pertaining to SIL character recognition systems. In the first step, the text is preprocessed and normalized to perform the text identification correctly. The second step involves extracting relevant and informative features. The third step implements the classification decision. The three stages which involved are Data acquisition and preprocessing, Feature extraction, and Classification. Here we concentrated on two techniques to obtain features, Feature Extraction & Feature Selection. Edge-hinge distribution is a feature that characterizes the changes in direction of a script stroke in handwritten text. The edge-hinge distribution is extracted by means of a windowpane that is slid over an edge-detected binary handwriting image. Whenever the mid pixel of the window is on, the two edge fragments (i.e. connected sequences of pixels) emerging from this mid pixel are measured. Their directions are measured and stored as pairs. A joint probability distribution is obtained from a large sample of such pairs. Despite continuous effort, handwriting identification remains a challenging issue, due to different approaches use different varieties of features, having different. Therefore, our study will focus on handwriting recognition based on feature selection to simplify features extracting task, optimize classification system complexity, reduce running time and improve the classification accuracy.

Keywords: word segmentation and recognition, character recognition, optical character recognition, hand written character recognition, South Indian languages

Procedia PDF Downloads 471

Authors: Tawfik Thelaidjia, Salah Chenikher

Abstract:

Bearing fault diagnosis has attracted significant attention over the past few decades. It consists of two major parts: vibration signal feature extraction and condition classification for the extracted features. In this paper, feature extraction from faulty bearing vibration signals is performed by a combination of the signal’s Kurtosis and features obtained through the preprocessing of the vibration signal samples using Db2 discrete wavelet transform at the fifth level of decomposition. In this way, a 7-dimensional vector of the vibration signal feature is obtained. After feature extraction from vibration signal, the support vector machine (SVM) was applied to automate the fault diagnosis procedure. To improve the classification accuracy for bearing fault prediction, particle swarm optimization (PSO) is employed to simultaneously optimize the SVM kernel function parameter and the penalty parameter. The results have shown feasibility and effectiveness of the proposed approach

Keywords: condition monitoring, discrete wavelet transform, fault diagnosis, kurtosis, machine learning, particle swarm optimization, roller bearing, rotating machines, support vector machine, vibration measurement

Procedia PDF Downloads 410

Authors: Yousif Mohamed Y. Abdallah, Razan Manofely, Rajab M. Ben Yousef

Abstract:

X-Ray images are very popular as a first tool for diagnosis. Automating the process of analysis of such images is important in order to help physician procedures. In this practice, teeth segmentation from the radiographic images and feature extraction are essential steps. The main objective of this study was to study correction preprocessing of x-rays images using local adaptive filters in order to evaluate contrast enhancement pattern in different x-rays images such as grey color and to evaluate the usage of new nonlinear approach for contrast enhancement of soft tissues in x-rays images. The data analyzed by using MatLab program to enhance the contrast within the soft tissues, the gray levels in both enhanced and unenhanced images and noise variance. The main techniques of enhancement used in this study were contrast enhancement filtering and deblurring images using the blind deconvolution algorithm. In this paper, prominent constraints are firstly preservation of image's overall look; secondly, preservation of the diagnostic content in the image and thirdly detection of small low contrast details in diagnostic content of the image.

Keywords: enhancement, x-rays, pixel intensity values, MatLab

Procedia PDF Downloads 446

Authors: Avatharam Ganivada, Krishna Shah

Abstract:

The Mask R-CNN is recently introduced by the team of Facebook AI Research (FAIR), which is mainly concerned with instance segmentation in images. Here, the Mask R-CNN is based on ResNet and feature pyramid network (FPN), where a single dropout method is employed. This paper provides a modified Mask R-CNN by adding multiple dropout methods into the Mask R-CNN. The proposed model has also utilized the concepts of Resnet and FPN to extract stage-wise network feature maps, wherein a top-down network path having lateral connections is used to obtain semantically strong features. The proposed model produces three outputs for each object in the image: class label, bounding box coordinates, and object mask. The performance of the proposed network is evaluated in the segmentation of every instance in images using COCO and cityscape datasets. The proposed model achieves better performance than the state-of-the-networks for the datasets.

Keywords: instance segmentation, object detection, convolutional neural networks, deep learning, computer vision

Procedia PDF Downloads 44

Authors: Ahd Aljarf, Saad Amin

Abstract:

Images are important source of information used as evidence during any investigation process. Their clarity and accuracy is essential and of the utmost importance for any investigation. Images are vulnerable to losing blocks and having noise added to them either after alteration or when the image was taken initially, therefore, having a high performance image processing system and it is implementation is very important in a forensic point of view. This paper focuses on improving the quality of the forensic images. For different reasons packets that store data can be affected, harmed or even lost because of noise. For example, sending the image through a wireless channel can cause loss of bits. These types of errors might give difficulties generally for the visual display quality of the forensic images. Two of the images problems: noise and losing blocks are covered. However, information which gets transmitted through any way of communication may suffer alteration from its original state or even lose important data due to the channel noise. Therefore, a developed system is introduced to improve the quality and clarity of the forensic images.

Keywords: image filtering, image reconstruction, image processing, forensic images

Procedia PDF Downloads 339

Authors: A. Degale Desta, Cheng Jian

Abstract:

Deep learning application in computer vision is rapidly advancing, giving it the ability to monitor the public and quickly identify potentially anomalous behaviour from crowd scenes. Therefore, the purpose of the current work is to improve the performance of safety of people in crowd events from panic behaviour through introducing the innovative idea of Aggregation of Ensembles (AOE), which makes use of the pre-trained ConvNets and a pool of classifiers to find anomalies in video data with packed scenes. According to the theory of algorithms that applied K-means, KNN, CNN, SVD, and Faster-CNN, YOLOv5 architectures learn different levels of semantic representation from crowd videos; the proposed approach leverages an ensemble of various fine-tuned convolutional neural networks (CNN), allowing for the extraction of enriched feature sets. In addition to the above algorithms, a long short-term memory neural network to forecast future feature values and a handmade feature that takes into consideration the peculiarities of the crowd to understand human behavior. On well-known datasets of panic situations, experiments are run to assess the effectiveness and precision of the suggested method. Results reveal that, compared to state-of-the-art methodologies, the system produces better and more promising results in terms of accuracy and processing speed.

Keywords: action recognition, computer vision, crowd detecting and tracking, deep learning

Procedia PDF Downloads 123

Authors: Liwei Liu, Shuguo Pan, Wang Gao

Abstract:

In order to take full advantages of the BDS-3 penta-frequency signals in the long-baseline RTK positioning, a long-baseline RTK positioning method based on the BDS-3 penta-frequency ionospheric-reduced (IR) combinations is proposed. First, the low noise and weak ionospheric delay characteristics of the multi-frequency combined observations of BDS-3is analyzed. Second, the multi-frequency extra-wide-lane (EWL)/ wide-lane (WL) combinations with long-wavelengths are constructed. Third, the fixed IR EWL combinations are used to constrain the IR WL, then constrain narrow-lane (NL)ambiguityies and start multi-epoch filtering. There is no need to consider the influence of ionospheric parameters in the third step. Compared with the estimated ionospheric model, the proposed method reduces the number of parameters by half, so it is suitable for the use of multi-frequency and multi-system real-time RTK. The results using real data show that the stepwise fixed model of the IR EWL/WL/NL combinations can realize long-baseline instantaneous cimeter-level positioning.

Keywords: penta-frequency, ionospheric-reduced (IR), RTK positioning, long-baseline

Procedia PDF Downloads 130

Authors: Baba Mbaye

Abstract:

In collaborative filtering recommendation systems, a user receives suggested items based on the opinions and evaluations of a community of users. This type of recommendation system uses only the information (notes in numerical values) contained in a usage matrix as input data. This matrix can be constructed based on users' behaviors or by offering users to declare their opinions on the items they know. The cold start problem leads to very poor performance for new users. It is a phenomenon that occurs at the beginning of use, in the situation where the system lacks data to make recommendations. There are three types of cold start problems: cold start for a new item, a new system, and a new user. We are interested in this article at the cold start for a new user. When the system welcomes a new user, the profile exists but does not have enough data, and its communities with other users profiles are still unknown. This leads to recommendations not adapted to the profile of the new user. In this paper, we propose an approach that improves cold start by using the notions of similarity and semantic proximity between users profiles during cold start. We will use the cold-metadata available (metadata extracted from the new user's data) useful in positioning the new user within a community. The aim is to look for similarities and semantic proximities with the old and current user profiles of the system. Proximity is represented by close concepts considered to belong to the same group, while similarity groups together elements that appear similar. Similarity and proximity are two close but not similar concepts. This similarity leads us to the construction of similarity which is based on: a) the concepts (properties, terms, instances) independent of ontology structure and, b) the simultaneous representation of the two concepts (relations, presence of terms in a document, simultaneous presence of the authorities). We propose an ontology, OIVCSRS (Ontology of Improvement Visitor Cold Start in Recommender Systems), in order to structure the terms and concepts representing the meaning of an information field, whether by the metadata of a namespace, or the elements of a knowledge domain. This approach allows us to automatically attach the new user to a user community, partially compensate for the data that was not initially provided and ultimately to associate a better first profile with the cold start. Thus, the aim of this paper is to propose an approach to improving cold start using semantic technologies.

Keywords: visitor cold start, recommender systems, collaborative filtering, semantic filtering

Procedia PDF Downloads 195

Authors: Kun Xu, Yuan Xu, Jia Qiao

Abstract:

The document detection plays an important role in optical character recognition and text analysis. Because the traditional detection methods have weak generalization ability, and deep neural network has complex structure and large number of parameters, which cannot be well applied in mobile devices, this paper proposes a lightweight Document Corner Detection Network (DCDNet). DCDNet is a two-stage architecture. The first stage with Encoder-Decoder structure adopts depthwise separable convolution to greatly reduce the network parameters. After introducing the Feature Attention Union (FAU) module, the second stage enhances the feature information of spatial and channel dim and adaptively adjusts the size of receptive field to enhance the feature expression ability of the model. Aiming at solving the problem of the large difference in the number of pixel distribution between corner and non-corner, Weighted Binary Cross Entropy Loss (WBCE Loss) is proposed to define corner detection problem as a classification problem to make the training process more efficient. In order to make up for the lack of Dataset of document corner detection, a Dataset containing 6620 images named Document Corner Detection Dataset (DCDD) is made. Experimental results show that the proposed method can obtain fast, stable and accurate detection results on DCDD.

Keywords: document detection, corner detection, attention mechanism, lightweight

Procedia PDF Downloads 326

Authors: M. Hilani, B. Nassih

Abstract:

Phishing is one of the oldest and best known scams on the Internet. It can be defined as any type of telecommunications fraud that uses social engineering tricks to obtain confidential data from its victims. It’s a cybercrime aimed at stealing your sensitive information. Phishing is generally done via private email, so scammers impersonate large companies or other trusted entities to encourage victims to voluntarily provide information such as login credentials or, worse yet, credit card numbers. The COVID-19 theme is used by cybercriminals in multiple malicious campaigns like phishing. In this environment, messaging filtering solutions have become essential to protect devices that will now be used outside of the secure perimeter. Despite constantly updating methods to avoid these cyberattacks, the end result is currently insufficient. Many researchers are looking for optimal solutions to filter phishing emails, but we still need good results. In this work, we concentrated on solving the problem of detecting phishing emails using the different steps of NLP preprocessing, and we proposed and trained a model using one-dimensional CNN. Our study results show that our model obtained an accuracy of 99.99%, which demonstrates how well our model is working.

Keywords: phishing, e-mail, NLP preprocessing, CNN, e-mail filtering

Procedia PDF Downloads 83

Authors: Peyman Sheikholharam Mashhadi

Abstract:

Difficulties in analyzing real world images in classical image processing and machine vision framework have motivated researchers towards considering the biology-based vision. It is a common belief that mammalian visual cortex has been adapted to the statistics of the real world images through the evolution process. There are two well-known successful models of mammalian visual cortical cells: Independent Component Analysis (ICA) and Independent Subspace Analysis (ISA). In this paper, we statistically analyze the dependencies which remain in the components after applying these models to the natural images. Also, we investigate the response of feature detectors to gratings with various parameters in order to find optimal parameters of the feature detectors. Finally, the selectiveness of feature detectors to phase, in both models is considered.

Keywords: statistics, independent component analysis, independent subspace analysis, phase, natural images

Procedia PDF Downloads 320

Authors: Adel Sabry Eesa, Adnan Mohsin Abdulazeez Brifcani, Zeynep Orman

Abstract:

Feature selection has been used in many fields such as classification, data mining and object recognition and proven to be effective for removing irrelevant and redundant features from the original data set. In this paper, a new design of distributed intrusion detection system using a combination feature selection model based on bees and decision tree. Bees algorithm is used as the search strategy to find the optimal subset of features, whereas decision tree is used as a judgment for the selected features. Both the produced features and the generated rules are used by Decision Making Mobile Agent to decide whether there is an attack or not in the networks. Decision Making Mobile Agent will migrate through the networks, moving from node to another, if it found that there is an attack on one of the nodes, it then alerts the user through User Interface Agent or takes some action through Action Mobile Agent. The KDD Cup 99 data set is used to test the effectiveness of the proposed system. The results show that even if only four features are used, the proposed system gives a better performance when it is compared with the obtained results using all 41 features.

Keywords: distributed intrusion detection system, mobile agent, feature selection, bees algorithm, decision tree

Procedia PDF Downloads 368

Authors: M. Amparo Núñez-Andrés, Felipe Buill, Albert Prades

Abstract:

More and more frequently, geomatics techniques such as terrestrial laser scanning or digital photogrammetry, either terrestrial or from drones, are being used to obtain digital terrain models (DTM) used for the monitoring of geological phenomena that cause natural disasters, such as landslides, rockfalls, debris-flow. One of the main multitemporal analyses developed from these models is the quantification of volume changes in the slopes and hillsides, either caused by erosion, fall, or land movement in the source area or sedimentation in the deposition zone. To carry out this task, it is necessary to filter the point clouds of all those elements that do not belong to the slopes. Among these elements, vegetation stands out as it is the one we find with the greatest presence and its constant change, both seasonal and daily, as it is affected by factors such as wind. One of the best-known indexes to detect vegetation on the image is the NVDI (Normalized Difference Vegetation Index), which is obtained from the combination of the infrared and red channels. Therefore it is necessary to have a multispectral camera. These cameras are generally of lower resolution than conventional RGB cameras, while their cost is much higher. Therefore we have to look for alternative indices based on RGB. In this communication, we present the results obtained in Georisk project (PID2019‐103974RB‐I00/MCIN/AEI/10.13039/501100011033) by using the GLI (Green Leaf Index) and ExG (Excessive Greenness), as well as the change to the Hue-Saturation-Value (HSV) color space being the H coordinate the one that gives us the most information for vegetation filtering. These filters are applied both to the images, creating binary masks to be used when applying the SfM algorithms, and to the point cloud obtained directly by the photogrammetric process without any previous filter or the one obtained by TLS (Terrestrial Laser Scanning). In this last case, we have also tried to work with a Riegl VZ400i sensor that allows the reception, as in the aerial LiDAR, of several returns of the signal. Information to be used for the classification on the point cloud. After applying all the techniques in different locations, the results show that the color-based filters allow correct filtering in those areas where the presence of shadows is not excessive and there is a contrast between the color of the slope lithology and the vegetation. As we have advanced in the case of using the HSV color space, it is the H coordinate that responds best for this filtering. Finally, the use of the various returns of the TLS signal allows filtering with some limitations.

Keywords: RGB index, TLS, photogrammetry, multispectral camera, point cloud

Procedia PDF Downloads 99

Authors: Hoda Ghavamipoor, S. Alireza Hashemi Golpayegani

Abstract:

An approach to model the customer interaction with e-commerce websites is presented. Considering the service quality level as a predictive feature, we offer an improved method based on the Customer Behavior Model Graph (CBMG), a state-transition graph model. To derive the Quality of Service sensitive-CBMG (QoS-CBMG) model, process-mining techniques is applied to pre-processed website server logs which are categorized as ‘buy’ or ‘visit’. Experimental results on an e-commerce website data confirmed that the proposed method outperforms CBMG based method.

Keywords: customer behavior model, electronic commerce, quality of service, customer behavior model graph, process mining

Procedia PDF Downloads 384

Authors: Zhang Shuqi, Liu Dan

Abstract:

For the problems in intelligent network anomaly traffic detection models, such as low detection accuracy caused by the lack of training samples, poor effect with small sample attack detection, a classification model enhancement method, F-ACGAN(Flow Auxiliary Classifier Generative Adversarial Network) which introduces generative adversarial network and adversarial training, is proposed to solve these problems. Generating adversarial data with category labels could enhance the training effect and improve classification accuracy and model robustness. FACGAN consists of three steps: feature preprocess, which includes data type conversion, dimensionality reduction and normalization, etc.; A generative adversarial network model with feature learning ability is designed, and the sample generation effect of the model is improved through adversarial iterations between generator and discriminator. The adversarial disturbance factor of the gradient direction of the classification model is added to improve the diversity and antagonism of generated data and to promote the model to learn from adversarial classification features. The experiment of constructing a classification model with the UNSW-NB15 dataset shows that with the enhancement of FACGAN on the basic model, the classification accuracy has improved by 8.09%, and the score of F1 has improved by 6.94%.

Keywords: data imbalance, GAN, ACGAN, anomaly detection, adversarial training, data augmentation

Procedia PDF Downloads 76