Browse Items (11808 total)
Sort by:
-
Financial Market Forecasting using Macro-Economic Variables and RNN
Stock market forecasting is widely recognized as one of the most important and difficult business challenges in time series forecasting. This is mainly due to its noise. The use of RNN algorithms for funding has attracted interest from traders and scientists. The best technique for learning long-term memory sequences is to use long and short networks. Based on the literature, it is acknowledged that LSTM neural networks outperform all other models. Macroeconomics is a discipline of economics that studies the behavior of the economy as a whole. Macroeconomic factors are economic, natural, geopolitical, or other variables that influence the economy of a country. This study studies and test several macroeconomic variables and their significance on stock market forecasting. In macroeconomics, we have series that are updated once a month or even once a quarter, with data that is rarely more than a few hundred characters long. The amount of data given can sometimes be insufficient for algorithms to uncover hidden patterns and generate meaningful results. Depending on the prediction needs, we proposed a feasible LSTM design and training algorithm. According to the findings of this study, the inclusion of macroeconomic variable has a significant impact on stock price prediction. 2022 IEEE. -
Road Accident Prediction using Machine Learning Approaches
Road accidents create a significant number of serious injuries reported per year and are a chief concern of the world, mostly in underdeveloped countries. Many people have lost their near and dear ones due to these road accidents. Hence a system that can potentially save lives is required. The system detects essential contributing elements for an accident or creates a link among accidents and various factors for the occurrence of accidents. This research proposes an Accident Prediction system that can help to analyze the potential safety issues and predict whether an accident will occur or not. A comparative study of various Machine Learning Algorithms was conducted to check which model can help predict accidents more accurately. The dataset used for this paper is the government record accidents that occurred in a district in India. Logistic Regression, Random Forest, Decision Tree, K-Nearest Neighbor, XGBoost, and Support Vector Machine are among the Machine Learning models used in this paper to predict accidents. The Random Forest algorithm gave the highest accuracy of 80.78% when the accuracies of the Machine Learning models were compared. 2022 IEEE. -
Cancer Tumor Detection Using Genetic Mutated Data and Machine Learning Models
Early detection of a disease is a crucial task because of unavailability of proper medical facilities. Cancer is one of the critical diseases that needs early detection for survival. A cancer tumor is caused due to thousands of genetic mutations. Understanding the genetic mutations of cancer tumor is a tedious and time-consuming task. A list of genetic variations is analysed manually by a molecular pathologist. The clinical strips of indication are of nine classes, but the classification is still unknown. The objective of this implementation is to suggest a multiclass classifier which classifies the genetic mutations with respect to the clinical signs. The clinical evidences are text-evidences of gene mutations and analysed by Natural Language Processing (NLP). Various machine learning concepts like Naive Bayes, Logistic Regression, Linear Support Vector Machine, Random Forest Classifier applied on the collected dataset which contain the evidence based on genetic mutations and other clinical evidences that pathology or specialists used to classify the gene mutations. The performances of the models are analysed to get the best results. The machine learning models are implemented and analyzed with the help of gene, variance and text features. Based on the variants of gene mutation, the risk of the cancer can be detected and the medications can be prescribed accordingly. 2022 IEEE. -
Deep Learning-based Gender Recognition Using Fusion of Texture Features from Gait Silhouettes
The gait of a person is the manner in which he or she walks. The human gait can be considered as a useful behavioral type of biometric that could be utilized for identifying people. Gait can also be used to identify a persons gender and age group. Recent breakthroughs in image processing and artificial intelligence have made it feasible to extract data from photographs and videos for various classifying purposes. Gender can be regarded as soft biometric that could be useful in video captured using surveillance cameras, particularly in uncontrolled environments with erratic placements. Gender recognition in security, particularly in surveillance systems, is becoming increasingly popular. Popularly used deep learning algorithms for images, convolutional neural networks, have proven to be a good mechanism for gender recognition. Still, there are drawbacks to convolutional neural network approaches, like a very complex network model, comparatively larger training time and highly expensive in computational resources, meager convergence quickness, overfitting of the network, and accuracy that may need improvement. As a result, this paper proposes a texture-based deep learning-based gender recognition system. The gait energy image, that is created by adding silhouettes received from a portion of the video which portrays an entire gait cycle, can be the most often utilized feature in gait-based categorization. More texture features, such as histogram of oriented gradient (HOG) and entropy for gender identification, have been examined in the proposed work. The accuracy of gender classification using whole body image, upper body image, and lower body image is compared in this research. Combining texture features is more accurate than looking at each texture feature separately, according to studies. Furthermore, full body gait images are more precise than partial body gait images. 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. -
Recommendation System using Clustering and Comparing Clustering and Topic Modelling Techniques
In this paper, we have used a technique called clustering to recommend the products to the customer and also tried to compare clustering and Topic modelling to find out which technique is better for our purpose. From all the papers that have been reviewed, we observed that the greater part of the proposal approaches applied content-based filtering (55%). Collaborative-based filtering was applied by just 18% of the looked into approaches, and hybrid based by 16%. Other suggestion ideas included generalizing, thing driven proposals, and crossover suggestions. The content-based filtering approaches overwhelmingly utilized papers that the clients had made, marked, examined, or downloaded [1]. To begin with, it stays muddled which suggestion ideas and approaches are the most encouraging. For instance, analysts demonstrated different results on the presentation of content based and collaborative filtering. A portion of the time content-based filtering performed better contrasted with collaborative filtering sand a portion of the time it performed all the more regrettable. 2022 IEEE. -
Political Optimizer Algorithm for Optimal Location and Sizing of Photovoltaic Distribution Generation in Electrical Distribution Network
In this paper, the political optimizer (PO), a new and efficient socio-inspired meta-heuristic search algorithm, is proposed for the first time in this research for determining the ideal locations and capacities of photovoltaic (PV) distribution generation (DG) in electrical distribution networks (EDN). A multi-objective function is designed to lower distribution losses and voltage deviation indexes and maximize voltage stability, among other objectives. The computational efficiency of PO when solving the optimal allocation of PV systems in EDN is investigated on an IEEE 33-bus EDN. The results indicate that integrating small DGs at multiple locations has a better EDN performance than integrating a single significant DG in the network. The results also suggest that, as demonstrated by a comparative analysis of PO results and those of other related literature works, PO can deal with complex multi-variable optimization problems. 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. -
Computer Assisted Unsupervised Extraction and Validation Technique for Brain Images from MRI
Magnetic Resonance Imaging (MRI) of human is a developing field in medical research because it assists in considering the brain anomalies. To identify and analyze brain anomalies, the research requires brain extraction. Brain extraction is a significant clinical image handling method for quick conclusion with clinical perception for quantitative assessment. Automated methods of extracting brain from MRI are challenging, due to the connected pixel intensity information for various regions such as skull, sub head and neck tissues. This paper presents a fully automated extraction of brain area from MRI. The steps involved in developing the method to extract brain area, includes image contrast limited using histogram, background suppression using average filtering, pixel region growing method by finding pixel intensity similarity and filling discontinuity inside brain region. Twenty volumes of brain slices are utilized in this research method. The outcome is achieved by this method is approved by comparing with manually extracted slices. The test results confirm the performance of this strategy can effectively section brain from MRI. 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. -
Malicious URL Detection Using Machine Learning Techniques
Cyber security is a very important requirement for users. With the rise in Internet usage in recent years, cyber security has become a serious concern for computer systems. When a user accesses a malicious Web site, it initiates a malicious behavior that has been pre-programmed. As a result, there are numerous methods for locating potentially hazardous URLs on the Internet. Traditionally, detection was based heavily on the usage of blacklists. Blacklists, on the other hand, are not exhaustive and cannot detect newly created harmful URLs. Recently, machine learning methods have received a lot of importance as a way to improve the majority of malicious URL detectors. The main goal of this research is to compile a list of significant features that can be utilized to detect and classify the majority of malicious URLs. To increase the effectiveness of classifiers for detecting malicious URLs, this study recommends utilizing host-based and lexical aspects of the URLs. Malicious and benign URLs were classified using machine learning classifiers such as AdaBoost and Random Forest algorithms. The experiment shows that Random Forest performs really well when checked using voting classifier on AdaBoost and Random Forest Algorithms. The Random Forest achieves about 99% accuracy. 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. -
A Study on Crude Oil Price Forecasting Using RNN Model
Crude oil forecasting plays an important role in every countrys economic progress. Inflation is likely to rise as oil prices rise, delaying economic progress. In terms of inflation, oil prices directly affect the expense of commodities produced using petroleum products. Not only crude, this paper provides the idea of best prediction models that could be used for easy prediction in stocks. It provides an overview of the data and methodology. As a result, we have compiled a list of articles that discuss the impact of crude oil on various stock markets and how it affects different countries. And in general, we were looking for the optimal price prediction model between gated recurrent units (GRUs) and long short-term memory (LSTM). 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. -
A Secure Communication Gateway with Parity Generator Implementation in QCA Platform
Quantum-Dot Cellular Automata (QCA) has arisen as a potential option in contrast to CMOS in the late time of nanotechnology. Some appealing highlights of QCA incorporate incredibly low force utilization and dissemination, high gadget pressing thickness, high velocity (arranged by THz). QCA based plans of normal advanced modules were concentrated broadly in the ongoing past. Equality generator and equality checker circuits assume a significant part in blunder discovery and subsequently, go about as fundamental segments in correspondence circuits. In any case, not very many endeavors were made for an efficient plan of QCA based equality generator as well as equality checker circuits up until now. In addition, these current plans need functional feasibility as they bargain a ton with normally acknowledged plan measurements like territory, postponement, intricacy, and manufacture cost. This article depicts new plans of equality generator and equality checker circuits in QCA which beat every one of the current plans as far as previously mentioned measurements. The proposed plans can likewise be effortlessly reached out to deal with an enormous number of contributions with a straight expansion in territory and inactivity. 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG. -
Artificial Intelligence Technological Revolution in Education and Space for Next Generation
The goal of this research is to discover the various potential for the educational system using artificial intelligence (AI). The world today is dealing with AI in different sectors. This study specifically looked into the prospects for acquiring efficient and high-quality education for each student, automating administrative tasks, including regulating adaptive student support systems. AI has been leveraged and used in the education sector in various formations. AI initially took in the form of computers with the cognitive model, transformed to online learning, together with other technologies, the use of AI provides chatbots to perform instructors. Imagine you can access your classroom from anywhere at any time through an online learning system. These functionalities enable the education system to deal with the curriculum effectively. Using these facilities, teachers instruct the students to desire to achieve their goals efficiently. The primary aim of this article addresses the concepts in AI that serve to regulate and improve the overall quality of academic performance. The secondary aim of this article is to discuss AI involvement in the space domain. 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. -
Secured Cloud Computing for Medical Database Monitoring Using Machine Learning Techniques
A growing number of people are calling on the health-care industry to adopt new technologies that are becoming accessible on the market in order to improve the overall quality of their services. Telecommunications systems are integrated with computers, connectivity, mobility, data storage, and information analytics to make a complete information infrastructure system. It is the order of the day to use technology that is based on the Internet of Things (IoT). Given the limited availability of human resources and infrastructure, it is becoming more vital to monitor chronic patients on an ongoing basis as their diseases deteriorate and become more severe. A cloud-based architecture that is capable of dealing with all of the issues stated above may be able to provide effective solutions for the health-care industry. With the purpose of building software that would mix cloud computing and mobile technologies for health-care monitoring systems, we have assigned ourselves the task of designing software. Using a method devised by Higuchi, it is possible to extract stable fractal values from electrocardiogram (ECG) data, something that has never been attempted previously by any other researcher working on the development of a computer-aided diagnosis system for arrhythmia. As a result of the results, it is feasible to infer that the support vector machine has attained the best classification accuracy attainable for fractal features. When compared to the other two classifiers, the feed forward neural network model and the feedback neural network model, the support vector machine excels them both. Furthermore, it should be noted that the sensitivity of both the feed forward neural network and the support vector machine yields results that are equivalent in quality (92.08% and 90.36%, respectively). 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. -
Linear Regression Tree and Homogenized Attention Recurrent Neural Network for Online Training Classification
Internet has become a vital part in people's life with the swift development of Information Technology (IT). Predominantly the customers share their opinions concerning numerous entities like, products, services on numerous platforms. These platforms comprises of valuable information concerning different types of domains ranging from commercial to political and social applications. Analysis of this immeasurable amount of data is both laborious and cumbersome to manipulate manually. In this work, a method called, Linear Regression Tree-based Homogenized Attention Recurrent Neural Network (LRT-HRNN) for online training is proposed. In the first step, a dataset consisting of student's reactions on E-learning is provided as input. A Linear Regression Decision Tree (LRT) - based feature (i.e., student's reactions and posts) selection model is applied in the second step. The feature selection model initially selects the commonly dispensed features. In the last step, HRNN sentiment analysis is employed for aggregating characterizations from prior and succeeding posts based on student's reactions for online training. During the experimentation process, LRT-HRNN method when compared with existing methods such as Attention Emotion-enhanced Convolutional Long Short Term Memory (AEC-LSTM) and Adaptive Particle Swarm Optimization based Long Short Term Memory (APSO-LSTM, performed better in terms of accuracy(increased by 6%), false positive rate (decreased by 22%), true positive rate (increased by 7%) and computational time (reduced by 21%). 2022 IEEE. -
Prevention of Data Breach by Machine Learning Techniques
In today's data communication environment, network and system security is vital. Hackers and intruders can gain unauthorized access to networks and online services, resulting in some successful attempts to knock down networks and web services. With the progress of security systems, new threats and countermeasures to these assaults emerge. Intrusion Detection Systems are one of these choices (IDS). An Intrusion Detection System's primary goal is to protect resources from attacks. It analyses and anticipates user behavior before determining if it is an assault or a common occurrence. We use Rough Set Theory (RST) and Gradient Boosting to identify network breaches (using the boost library). When packets are intercepted from the network, RST is used to pre-process the data and reduce the dimensions. A gradient boosting model will be used to learn and evaluate the features chosen by RST. RST-Gradient boost model provides the greatest results and accuracy when compared to other scale-down strategies like regular scaler. 2022 IEEE. -
A Comparison of 2 Step Classification with 3-Class Classification for Webpage Classification
The content over internet increasing significantly each year and the web page classification is an essential areas of work upon for web-based information management, content retrieval, data scrapping, content filtering, advertisement removal, contextual advertising, expanding web directories etc. Multiclass classification methods is more popular and commonly use to classify web pages, and 2 step classification is our proposed system. In 2 step classifier, we use 2 primary model which works serially and perform binary classification at each level. The primary source of dataset contain thousands of URL(Uniform Resource Locator) of web pages. The content on webpage is extracted and stored on system to avoid the loss of data sue to the change in URL. The comparison between the two methods validated the system improvement and improved in different metric such as precision and recall using 2 step classification technique. 2 step classification technique is faster while training and also shows performance improvement. There proposed system shows improvement in the performance of the results but not something significant. 2022 IEEE. -
Inverse Hilbert Fractal-Metamaterial Rings for Microstrip Antennas and Wideband Applications
A Novel Metamaterial (MTM) property is obtained using a fractal pattern known as Inverse Hilbert. The Mu-negative(MNG) characteristics have been recovered by adopting NRW method. This MTM characteristic is studied for 2.45 GHz using FR4 epoxy as substrate. The dimension of the substrate is 30mm36mm 1.6mm. This fractal metamaterial structure can be amalgamated with an optimized Microstrip antenna (MSA) for improvement in antenna parameters and can be used for RF energy harvesting. 2022 IEEE. -
Stock Price Prediction using Deep Learning and FLASK
The forecasting of stock prices is one of the most explored issues, and it attracts the attention of both academics and business professionals. It is quite difficult to make predictions about the stock market, and it takes extensive research into the patterns of data. With the expansion of the internet and indeed the growth of social media, online media and opinions frequently mirror investor sentiment. The volatility and non-linear structure of the financial stock markets makes accurate forecasting difficult. One of the sophisticated analysis techniques that is being used by academics in a variety of fields is the neural network. In this paper, we proposed deep learning techniques for google stock price prediction. A dataset from Kaggle was collected and applied deep learning techniques RNN, LSTM variants. We achieved better results with Bidirectional LSTM. We also created a web app for stock prediction using Christ University python FLASK. 2022 IEEE. -
Content Based Deep Factorization Framework for Scientific Article Recommender System
With the advancement in technology and the tremendous number of citations available in the digital libraries, it has become difficult for the research scholars to find a relevant set of reference papers. The accelerating rate of scientific publications results in the problem of information overload because of which the scholars spend their 70% of the time finding relevant papers. A citation recommendation system resolves the issue of spending a good amount of time and other resources for collecting a set of papers by providing the user with personalised recommendations of the articles. Existing state of art models do not take high-low order feature interactions into consideration, due to which the recommendations are not up to the desired level of performance. In this paper, we propose a content-based model which combines Deep Neural Network (DNN) and Factorization Machines (FM) where no pre-trainings are required for providing the citation recommendations. 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. -
Forecasting Prices of Black Pepper in Kerala and Karnataka using Univariate and Multivariate Recurrent Neural Networks
Our country has a high level of agricultural employment. Price swings harm the economy of our country. To combat this impact, forecasting the selling price of agricultural products has become a need. Forecasts of agricultural prices assist farmers, government officials, businesses, central banks, policymakers, and consumers. Price prediction can then assist in making better selections in this area. Black pepper, sometimes known as the "King of Spices, " is a popular spice farmed and exported in India. The largest producers of black pepper are Karnataka and Kerala. For black pepper in Kerala and Karnataka, this study provides a univariate and multivariate price prediction model using Recurrent Neural Networks (RNN), Long Short-Term Memory (LSTM), and Gated Recurrent Unit (GRU). The data is denoised using Singular Spectral Analysis (SSA). The most accurate method is the multivariate variate LSTM technique, which uses macroeconomic variables. It has a Mean Absolute Percentage Error (MAPE) of 0.012 and 0.040 for Kerala and Karnataka, respectively. Grenze Scientific Society, 2022. -
Efficient Routing Strategies for Energy Management in Wireless Sensor Network
Wireless Sensor Network (WSN) refers to a group of distributed sensors that are used to examine and record the physical circumstances of the environment and coordinate the collected data at the centre of the location. This WSN plays a significant role in providing the needs of routing protocols. One of the important aspects of routing protocol in accordance with Wireless Sensor Network is that they should be efficient in the consumption of energy and have a prolonged life for the network. In modern times, routing protocol, which is efficient in energy consumption, is used for Wireless Sensor Network. The routing protocol that is efficient in energy consumption is categorized into four main steps: CM Communication Model, Reliable Routing, Topology-Based Routing and NS Network Structure. The network structure can be further classified as flat/hierarchical. The communication model can be further classified as query, coherent/non-coherent, negotiation-based routing protocol system. The topology-based protocol can be further classified as mobile or location-based. Reliable routing can be further classified as QoS (Quality of service) or multiple-path based. A survey on routing protocol that is energy-efficient on Wireless Sensor Network has also been provided in this research. The Author(s), under exclusive license to Springer Nature Switzerland AG 2022.