scholarly journals Classification and Analysis of Malicious Traffic with Multi-layer Perceptron Model

2021 ◽  
Vol 26 (3) ◽  
pp. 303-310
Author(s):  
Shilpa P. Khedkar ◽  
Aroul Canessane Ramalingam
Keyword(s):  
Machine Learning Algorithms ◽  
Traffic Classification ◽  
Multi Layer Perceptron ◽  
Qos Provisioning ◽  
Management Information ◽  
The Past ◽  
Information Classification ◽  
Important Field

Traffic classification is very important field of computer science as it provides network management information. Classification of traffic become complicated due to emerging technologies and applications. It is used for Quality of Service (QoS) provisioning, security and detecting intrusion in a system. In the past used of port, inspecting packet, and machine learning algorithms have been used widely, but due to the sudden changes in the traffic, their accuracy was diminished. In this paper a Multi-Layer Perceptron model with 2 hidden layers is proposed for traffic classification and target traffic classify into different categories. The experimental results indicate that proposed classifier efficiently classifies traffic and achieves 99.28% accuracy without feature engineering.

scholarly journals Meta-Learner for Amharic Sentiment Classification

2021 ◽  
Vol 11 (18) ◽  
pp. 8489
Author(s):  
Girma Neshir ◽  
Andreas Rauber ◽  
Solomon Atnafu
Keyword(s):  
Sentiment Analysis ◽  
Performance Indicator ◽  
Cost Effective ◽  
Machine Learning Algorithms ◽  
Sentiment Classification ◽  
Training Data ◽  
Policy Makers ◽  
The Past ◽  
Classification Framework

The emergence of the World Wide Web facilitates the growth of user-generated texts in less-resourced languages. Sentiment analysis of these texts may serve as a key performance indicator of the quality of services delivered by companies and government institutions. The presence of user-generated texts is an opportunity for assisting managers and policy-makers. These texts are used to improve performance and increase the level of customers’ satisfaction. Because of this potential, sentiment analysis has been widely researched in the past few years. A plethora of approaches and tools have been developed—albeit predominantly for well-resourced languages such as English. Resources for less-resourced languages such as, in this paper, Amharic, are much less developed. As a result, it requires cost-effective approaches and massive amounts of annotated training data, calling for different approaches to be applied. This research investigates the performance of a combination of heterogeneous machine learning algorithms (base learners such as SVM, RF, and NB). These models in the framework are fused by a meta-learner (in this case, logistic regression) for Amharic sentiment classification. An annotated corpus is provided for evaluation of the classification framework. The proposed stacked approach applying SMOTE on TF-IDF characters (1,7) grams features has achieved an accuracy of 90%. The overall results of the meta-learner (i.e., stack ensemble) have revealed performance rise over the base learners with TF-IDF character n-grams.

Automatic Segmentation and Classification of Outdoor Images Using Neural Networks

1997 ◽  
Vol 08 (01) ◽  
pp. 137-144 ◽  
Author(s):  
N. W. Campbell ◽  
B. T. Thomas ◽  
T. Troscianko
Keyword(s):  
Neural Networks ◽  
Automatic Segmentation ◽  
Texture Features ◽  
Multi Layer Perceptron ◽  
Feature Map ◽  
Segmentation Process ◽  
Image Area

The paper describes how neural networks may be used to segment and label objects in images. A self-organising feature map is used for the segmentation phase, and we quantify the quality of the segmentations produced as well as the contribution made by colour and texture features. A multi-layer perceptron is trained to label the regions produced by the segmentation process. It is shown that 91.1% of the image area is correctly classified into one of eleven categories which include cars, houses, fences, roads, vegetation and sky.

scholarly journals A Review : Classification of Food Grains and Quality Prediction

2020 ◽  
pp. 34-37
Author(s):  
Mahalaxmi P P ◽  
Kavita D. Hanabaratti
Keyword(s):  
Machine Learning ◽  
Learning Algorithms ◽  
Machine Learning Algorithms ◽  
Automated System ◽  
Quality Prediction ◽  
Food Grains ◽  
Geometrical Features ◽  
Image Procession

This review paper discuss about recent techniques and methods used for grain classification and grading. Grains are important source of nutrients and they play important role in healthy diet. The production of grains across worldwide each year is in terms of hundreds of millions. The common method to classify these hugely produced grains is manual which is mind-numbing and not accurate. So the automated system is required which can classify the verities and predict the quality (i.e. grade A, grade B) of grain fast and accurate. As machine learning had done most of the difficult things easy, many machine learning algorithms can be used which can easily classify and predict the quality of grains. The system uses colour and geometrical features like size and area of grains as attributes for classification and quality prediction. Here, several image procession methods and machine learning algorithms are reviewed.

scholarly journals A Smart Sensing Architecture for Domestic Monitoring: Methodological Approach and Experimental Validation

Sensors ◽  
2018 ◽  
Vol 18 (7) ◽  
pp. 2310 ◽  
Author(s):  
Andrea Monteriù ◽  
Mario Prist ◽  
Emanuele Frontoni ◽  
Sauro Longhi ◽  
Filippo Pietroni ◽  
...  
Keyword(s):  
Sensor Network ◽  
Life Quality ◽  
Well Being ◽  
Heterogeneous Data ◽  
Machine Learning Algorithms ◽  
Daily Routine ◽  
Smart Sensing ◽  
At Home

Smart homes play a strategic role for improving life quality of people, enabling to monitor people at home with numerous intelligent devices. Sensors can be installed to provide a continuous assistance without limiting the resident’s daily routine, giving her/him greater comfort, well-being and safety. This paper is based on the development of domestic technological solutions to improve the life quality of citizens and monitor the users and the domestic environment, based on features extracted from the collected data. The proposed smart sensing architecture is based on an integrated sensor network to monitor the user and the environment to derive information about the user’s behavior and her/his health status. The proposed platform includes biomedical, wearable, and unobtrusive sensors for monitoring user’s physiological parameters and home automation sensors to obtain information about her/his environment. The sensor network stores the heterogeneous data both locally and remotely in Cloud, where machine learning algorithms and data mining strategies are used for user behavior identification, classification of user health conditions, classification of the smart home profile, and data analytics to implement services for the community. The proposed solution has been experimentally tested in a pilot study based on the development of both sensors and services for elderly users at home.

scholarly journals Semi-automatic decision-making process in histopathological specimens from Barrett’s carcinoma patients using hyperspectral imaging (HSI)

2020 ◽  
Vol 6 (3) ◽  
pp. 261-263
Author(s):  
Marianne Maktabi ◽  
Hannes Köhler ◽  
Claire Chalopin ◽  
Thomas Neumuth ◽  
Yannis Wichmann ◽  
...  
Keyword(s):  
Hyperspectral Imaging ◽  
Near Infrared ◽  
Cancer Diagnostics ◽  
Machine Learning Algorithms ◽  
Infrared Range ◽  
Decision Making Process ◽  
Multi Layer Perceptron ◽  
Classification Methods ◽  
Near Infrared Range

AbstractDiscrimination of malignant and non-malignant cells of histopathologic specimens is a key step in cancer diagnostics. Hyperspectral Imaging (HSI) allows the acquisition of spectra in the visual and near-infrared range (500-1000nm). HSI can support the identification and classification of cancer cells using machine learning algorithms. In this work, we tested four classification methods on histopathological slides of esophageal adenocarcinoma. The best results were achieved with a Multi-Layer Perceptron. Sensitivity and F1-Score values of 90% were obtained.

scholarly journals Development of building information classifier elements to create and maintain information models of capital construction objects in terms of design processes, construction process management and construction information

Vestnik MGSU ◽  
2021 ◽  
pp. 926-954
Author(s):  
Vladimir S. Timchenko ◽  
Vladimir A. Volkodav ◽  
Ivan A. Volkodav ◽  
Olga V. Timchenko ◽  
Nikita A. Osipov
Keyword(s):  
Process Management ◽  
Design Processes ◽  
Information Classification ◽  
Information Models ◽  
Management Processes ◽  
Building Information ◽  
Capital Construction ◽  
The Cost

Introduction. Integral approach to the application of construction information in the creation and maintenance of information models of capital construction objects is key in the constant development of construction activities. Besides, according to the global trends, the direct implementation of construction activity including construction of especially complicated and unique objects and typification of classical ones requires application of a unified system of building information classification to optimize duration, costs and improve the quality of the constructed object. Development of the Russian classifier of building information was the first step in this direction allowing to make a tool which is the unified system of building information classification generally available. The Russian building information classifier developed in 2020 contains a lot of elements among which we can distinguish those groups which allow to manage the cost, duration and quality of the future capital construction object both at the early stages of its life cycle and later: management processes, design processes and information. Materials and methods. International systems of classification of building information that have found wide practical application in the field of construction: OmniClass (USA), Uniclass 2015 (Great Britain), CCS (Denmark) and CoClass (Sweden) are considered. The analysis of the structures and composition of existing classification systems, as well as the analysis of the current regulatory and technical framework in the Russian Federation in the field of construction in areas related to the management of processes, design of capital construction object and its information entities. Results. Taking into account the analysis and generalization of world practice in the field of construction, and classification of building information, parts of the building information classifier adapted to the specifics of the national base of normative and technical documentation in construction, applicable to the design and management of capital construction object, as well as for its description, were developed. The structure recommended by the standard ISO 12006-2:2015 is adopted as the basis for such classification tables of the building information classifier. When developing the composition of the classifier, the requirements for unification and standardization of existing national classifiers and experience in the construction industry on domestic and foreign objects were taken into account. Classification tables of the building information classifier for the two areas of activity in construction (Management, Design) and a classification table describing the information entities of the capital construction object were developed. Conclusions. Classification tables “Process Management”, “Design Processes”, “Information” of the building information classifier in the developed structures and composition provide the formation of a unified structure of management and design of capital construction object, allowing to combine its parts for adaptation to the requirements of a particular object and organization. Thus, providing an opportunity to optimize its technical and economic indicators, including the duration of construction and the cost of the object in the extent of its life cycle, to develop a tool for typing design and management processes, including planning tools and quality and cost control. An additional tool for the interrelation of various activities in construction (e.g., design, operation, construction, etc.) is the developed classification table “Information”, which describes the information entities of the capital construction object.

scholarly journals Culture-Bound Web-Based Institutional Academic Texts. The Case of Some Romance Language University Websites in English

Linguaculture ◽  
2019 ◽  
Vol 10 (1) ◽  
pp. 113-131
Author(s):  
Ioana-Carmen Păștinaru
Keyword(s):  
Web Based ◽  
Translation Strategies ◽  
Romance Language ◽  
University Websites ◽  
The Past ◽  
Academic Texts ◽  
Study Results ◽  
The Impact

The internationalisation process of European higher education over the past years largely encouraged the translation into English of many university websites. However, the (deliberate or nondeliberate) presence of culture-bound terms on the English version of university websites represents an issue of debate, considering the worldwide provenance of visitors accessing the websites and the purpose of these texts. The main goal of this article is to analyse the appropriateness of translation strategies used for the culture-bound terms on university websites. The practical part of this research uses Aixelá’s classification of translation strategies for the analysis of the culture-bound terms identified on some Romance language university webpages translated into English, allowing a series of suggestions and recommendations in each case. The study results have demonstrated that the strategy of conservation through repetition is used most often. Last but not least, this paper intends to raise awareness as to the translator’s role and the impact of the quality of translations of university webpages into English as a lingua franca.

scholarly journals AUV localisation: a review of passive and active techniques

2021 ◽  
Author(s):  
Francesco Maurelli ◽  
Szymon Krupiński ◽  
Xianbo Xiang ◽  
Yvan Petillot
Keyword(s):  
Underwater Vehicles ◽  
Machine Learning Algorithms ◽  
Mobile Systems ◽  
Robotic Systems ◽  
High Demand ◽  
Rapid Improvement ◽  
The Past ◽  
Challenging Environment ◽  
Water Blocking

AbstractLocalisation, i.e. estimation of one’s position in a given environment is a crucial element of many mobile systems, manned and unmanned. Due to the high demand for autonomous exploration, patrolling and inspection services and a rapid improvement of batteries, sensors and machine learning algorithms, the quality of localisation becomes even more important for smart robotic systems. The underwater domain is a very challenging environment due to the water blocking most of the signals over short distances. Recent results in localisation techniques for underwater vehicles are summarised in two principal categories: passive techniques, which strive to provide the best estimation of the vehicle’s position (global or local) given the past and current information from sensors, and active techniques, which additionally produce guidance output that is expected to minimise the uncertainty of estimated position.

scholarly journals A data-driven methodology to discover similarities between cocaine samples

2020 ◽  
Vol 10 (1) ◽  
Author(s):  
Fidelia Cascini ◽  
Nadia De Giovanni ◽  
Ilaria Inserra ◽  
Federico Santaroni ◽  
Luigi Laura
Keyword(s):  
Machine Learning ◽  
Machine Learning Algorithms ◽  
Main Hypothesis ◽  
Web Based ◽  
Science Field ◽  
A Value ◽  
Similar Drug ◽  
Web Based System

Abstract Machine learning has been used for distinct purposes in the science field but no applications on illegal drug have been done before. This study proposes a new web-based system for cocaine classification, profiling relations and comparison, that is capable of producing meaningful output based on a large amount of chemical profiling’s data. In particular, the Profiling Relations In Drug trafficking in Europe (PRIDE) system, offers several advantages to intelligence actions across Europe. Thus, it provides a standardized, broad methodology which uses machine learning algorithms to classify and compare drug profiles, highlight how similar drug samples are, and how probable it is that they share a common origin, batch, or preparation process. We evaluated the proposed algorithms using precision and recall metrics and analyzed the quality of predictions performed by the algorithms, with respect to our gold standard. In our experiments, we reached a value of 88% for F0.5-measure, 91% for precision, and 78% for recall, confirming our main hypothesis: machine learning can learn and be applied to have an automatic classification of cocaine profiles.

scholarly journals A Two-Branch CNN Architecture for Land Cover Classification of PAN and MS Imagery

2018 ◽  
Vol 10 (11) ◽  
pp. 1746 ◽  
Author(s):  
Raffaele Gaetano ◽  
Dino Ienco ◽  
Kenji Ose ◽  
Remi Cresson
Keyword(s):  
Deep Learning ◽  
Land Cover ◽  
Land Cover Classification ◽  
Machine Learning Algorithms ◽  
Learning Approaches ◽  
Sensing Applications ◽  
Very High Spatial Resolution ◽  
Urban Monitoring

The use of Very High Spatial Resolution (VHSR) imagery in remote sensing applications is nowadays a current practice whenever fine-scale monitoring of the earth’s surface is concerned. VHSR Land Cover classification, in particular, is currently a well-established tool to support decisions in several domains, including urban monitoring, agriculture, biodiversity, and environmental assessment. Additionally, land cover classification can be employed to annotate VHSR imagery with the aim of retrieving spatial statistics or areas with similar land cover. Modern VHSR sensors provide data at multiple spatial and spectral resolutions, most commonly as a couple of a higher-resolution single-band panchromatic (PAN) and a coarser multispectral (MS) imagery. In the typical land cover classification workflow, the multi-resolution input is preprocessed to generate a single multispectral image at the highest resolution available by means of a pan-sharpening process. Recently, deep learning approaches have shown the advantages of avoiding data preprocessing by letting machine learning algorithms automatically transform input data to best fit the classification task. Following this rationale, we here propose a new deep learning architecture to jointly use PAN and MS imagery for a direct classification without any prior image sharpening or resampling process. Our method, namely M u l t i R e s o L C C , consists of a two-branch end-to-end network which extracts features from each source at their native resolution and lately combine them to perform land cover classification at the PAN resolution. Experiments are carried out on two real-world scenarios over large areas with contrasted land cover characteristics. The experimental results underline the quality of our method while the characteristics of the proposed scenarios underline the applicability and the generality of our strategy in operational settings.

Sign in / Sign up

Export Citation Format

Share Document