Lahore Garrison University Research Journal of Computer Science and Information Technology
Not a member yet
227 research outputs found
Sort by
Trading Algorithm Model Based on Technical Indicators
No full textToday the rapid proliferation of the internet provides an environment where efficient e-commerce solutions can be developed. The electronic market is gaining more attention in the global economy, it gives buyers and sellers more liberty to trade cost-effectively and allows access to an adequate amount of data for analysis. New trading agents have been developed for the best utilization of such data. These agents design strategies using financial analysis techniques such as technical indicators. Two very well-known technical indicators used to develop strategies are Convergence-Divergence (MACD) and Stochastic Oscillator (SO). This paper aims to devise a trading algorithm that combines MACD and SO in a single strategy and check the reliability of the combined signals it generates. JTAP simulation system has been used to test the proposed strategy. In this paper, we evaluated the performance of our proposed strategy when implemented on shares of Karachi Stock Exchange, Pakistan which proves improvement of strategy
ROLE OF MACHINE VISION FOR IDENTIFICATION OF KIDNEY STONES USING MULTI FEATURES ANALYSIS
Get PDFThe purpose of this study is to highlight the significance of machine vision for the Classification of kidney stone identification. A novel optimized fused texture features frame work was designed to identify the stones in kidney. A fused 234 texture feature namely (GLCM, RLM and Histogram) feature set was acquired by each region of interest (ROI). It was observed that on each image 8 ROI’s of sizes (16x16, 20x20 and 22x22) were taken. It was difficult to handle a large feature space 280800 (1200x234). Now to overcome this data handling issue we have applied feature optimization technique namely POE+ACC and acquired 30 most optimized features set for each ROI. The optimized fused features data set 3600(1200x30) was used to four machine vision Classifiers that is Random Forest, MLP, j48 and Naïve Bayes. Finally, it was observed that Random Forest provides best results of 90% accuracy on ROI 22x22 among the above discussed deployed Classifier
Hybrid Image Steganography Method with Random Embedding of Encrypted Message
Get PDFThe main challenge for embedding encrypted message in an input image is to get better the security of the confidential information through hybrid-based image steganography method. Moreover, earlier LSB based solutions existed in which either secret information embedded without encryption or embedded un-randomly in an image and existing MSB based information concealing solutions minimizes information capacity and image quality too. Most of existing steganographic systems either based on LSB or MSB but only some hybrid solutions are available in which either the confidential message is not encoded before embedding it into the image and the embedding system is also not random based. The existing well known hybrid based image steganography techniques are not only deficient in performance but also deficient in embedding of encoded data in an image. To overcome these issues, a Hybrid-LSB-MSB based image steganography and multi-operation data encryption method is proposed in this article. Proposed method is not only randomly embeds the confidential information in a cover image but also provided the facility to encode the confidential information before substituting. The Hybrid-LSB-MSB based proposed image steganography method is compared with earlier Hybrid based image steganography method by using Mean Squared Error (MSE) and Peak Signal to Noise Ratio (PSNR) values including payload capacity. Higher PSNR and Lower MSE values signify effective steganography quality. The experimental results show that proposed method retains higher PSNR and lesser MSE values as contrasted to the existing methods thereby effective in steganographic properties.  
Detection of Crime Patterns in Digital Forensic Investigation to Trace the Adversaries
No full textThe use of the internet has increased significantly over the past couple of years. Access to the internet has become so common that a person without computer knowledge can also use this facility easily. This ease of availability has provided a lot of benefits to society but on the other hand misuse of the internet for personal or corporate benefits is also increasing. To prosecute cybercriminals and make some lawful checks on everyone's digital activities, digital forensic science comes into the light. In this context, we developed a new framework that improves the digital forensic investigation process. This research paper proposes a method in which we can identify the illegal activities and trace the adversaries. We capture the TCP (Transmission Control Protocol) packets from the servers and workstations. This data collected from the TCP log is stored in the database and preprocessed to eliminate redundant data. Furthermore, the database also contains past data. The proposed framework has three major processes collection of TCP packets, storing and preprocessing of collected data in a database, and mining of the pattern through a digital forensic anomaly collection algorithm. For the evaluation of our proposed framework, we have developed a java based application. The results are shown in the form of reports and tables
A Survey on Emotion Detection from Text in Social Media Platforms
No full textThis paper provides an overview of the evolving field of emotion detection and identifies the current generation of methods of emotion detection from social media platforms as well as the challenges. The challenges in the field of current emotion detection are discussed in detail and potential alternatives are proposed to enhance the ability to detect emotions in real-life systems that emphasize interactions between humans and computers as well as advertisements, recommendation systems, and medical fields such as computer-based therapy. These solutions include the extraction of semantic analysis keywords, and ontology design with the evaluation of emotions. There are multiple models and classifications of emotions such as Ekman’s model (Happy, Anger, Sad, Disgust,Fear, Surprise), and Plutchik’s model (anger-fear, surprise-anticipation, joy-sadness, joy-sadness). Further, a systematic review of publications on textual emotions detection from social media platforms, state-of-the-art methods, and existing challenges presented. Finally, we conclude with some recommendations based on critical analysis of existing techniques and determine future research directions presented at last
Classical and Probabilistic Information Retrieval Techniques: An Audit
Get PDFInformation retrieval is acquiring particular information from large resources and presenting it according to the user’s need. The incredible increase in information resources on the Internet formulates the information retrieval procedure, a monotonous and complicated task for users. Due to over access of information, better methodology is required to retrieve the most appropriate information from different sources. The most important information retrieval methods include the probabilistic, fuzzy set, vector space, and boolean models. Each of these models usually are used for evaluating the connection between the question and the retrievable documents. These methods are based on the keyword and use lists of keywords to evaluate the information material. In this paper, we present a survey of these models so that their working methodology and limitations are discussed. This is an important understanding because it makes possible to select an information retrieval technique based on the basic requirements. The survey results showed that the existing model for knowledge recovery is somewhere short of what was planned. We have also discussed different areas of IR application where these models could be used
Cloud Computing: Needs Enabling Data Mining and Business Intelligent Applications
No full textAs a new computational paradigm, cloud computing is attracting a lot of interest from researchers in the field of the business community and information technology sector and it can integrate with several heterogeneous resources makes it distinguished and unique to fulfill the demands of different types of users. The rapid increase in data volume and fixed access to online resources, which is related to all departments need to mine data for the discovery of knowledge. Its principal peculiarities incorporate a versatile asset design and along this line a suitable system for tending to be comprehended in an ideal mode. From the specific situations where cloud computing can be integrated, its use in business information and intelligence also conveys the highest aspirations from data mining to updates. This study gives an outline of the recent condition of the arrangement of Cloud Computing and elaborates, its implications in Business Intelligence and Data Mining. This study defines multiple layers that are expected to create such a framework in distinctive levels of deliberation, from the fundamental equipment stages to the product assets accessible to actualize the applications. At the end of this study, a few cases related to Data mining methodologies have been relocated to the Cloud Computing paradigm
Load Balancing in Cloud Computing Empowered with Dynamic Divisible Load Scheduling Method
Get PDFThe need to process and dealing with a vast amount of data is increasing with the developing technology. One of the leading promising technology is Cloud Computing, enabling one to accomplish desired goals, leading to performance enhancement. Cloud Computing comes into play with the debate on the growing requirements of data capabilities and storage capacities. Not every organization has the financial resources, infrastructure & human capital, but Cloud Computing offers an affordable infrastructure based on availability, scalability, and cost-efficiency. The Cloud can provide services to clients on-demand, making it the most adapted system for virtual storage, but still, it has some issues not adequately addressed and resolved. One of those issues is that load balancing is a primary challenge, and it is required to balance the traffic on every peer adequately rather than overloading an individual node. This paper provides an intelligent workload management algorithm, which systematically balances traffic and homogeneously allocates the load on every node & prevents overloading, and increases the response time for maximum performance enhancement
Weed Identification Methodology by using Transfer Learning
No full textFrom recent past years, Weed identification remained a hot topic for researchers. Majority of work focused on the detection of weed but we are trying to identify the weed via weed name. The unrivaled successes of deep learning make the researchers able to evaluate different weed species in the complex rangeland climate. Nowadays, with an increasing population, farming productivity needs to be increased a lot to meet the demand for accurate weed detection. Increased demand for an increase in the use of herbicides, resulting in environmental harm. In this research work, the picture of weed helps to detect and differentiate as per area, and its name. The main aim of this research is the identification of weed so that fewer herbicides can use. This research work will contribute toreducing the higher use of herbicides by helping clear identification of weed names through its features. We use transfer learning in machine learning. The deep Weeds dataset is used for the evaluation. For this, we use the deep learning model ResNet50 to get better results. The Deep Weeds dataset contains 17,509 images that are label and eight nationally recognized species of weed belonged to 8 across northern Australia locations. This paper declares a baseline for classification performance on the dataset of weed while utilizing the deep learning model ResNet-50 and it is a benchmark too. Deep learning model ResNet-50 attained an average accuracy classification of 96.16. The findings are high enough to make effective use of weed control methods in Pakistan for futurefield implementation. The results confirm that our System offers more effective Weed recognition than many other systems
Time Dependent Popularity Caching Scheme for NDN Based MANETs
No full textNamed data networking (NDN) approach has natural benefits within Mobile Ad hoc Network (MANET) but presents different issues as well. Space for cache, energy, and mobility of devices in a MANET is limited; therefore, we need for an enhanced judgement concerning which data to be store and where to be cache. A Time dependent Popularity Caching Scheme (TDPC) has suggested which selects nodes for caching the content on the forwarding path of packet and chooses the contents which have cached constructed on their time dependent popularity. At this interval, the cache distribution of the content and the storage capability of the devices are also measured. Results of the suggested TDPC approach are evaluated by using the simulator ndnSIM which is beached on Network Simulator 3 (NS-3). Simulation outcomes show that TDPC has good performance in expression of cache hit ratio, content retrieval interval, total cache copies and compared to the Dynamic Caching Strategy for CCN-based MANETs (CSCM). The goal of TDPC is to reduce cache redundancy, retrieval time of content and total number of cache copies