Commenced in January 2007
Frequency: Monthly
Edition: International
Paper Count: 258

World Academy of Science, Engineering and Technology

[Computer and Systems Engineering]

Online ISSN : 1307-6892

258 Data Clustering Algorithm Based on Multi-Objective Periodic Bacterial Foraging Optimization with Two Learning Archives

Authors: Chen Guo, Heng Tang, Ben Niu


Clustering splits objects into different groups based on similarity, making the objects have higher similarity in the same group and lower similarity in different groups. Thus, clustering can be treated as an optimization problem to maximize the intra-cluster similarity or inter-cluster dissimilarity. In real-world applications, the datasets often have some complex characteristics: sparse, overlap, high dimensionality, etc. When facing these datasets, simultaneously optimizing two or more objectives can obtain better clustering results than optimizing one objective. However, except for the objectives weighting methods, traditional clustering approaches have difficulty in solving multi-objective data clustering problems. Due to this, evolutionary multi-objective optimization algorithms are investigated by researchers to optimize multiple clustering objectives. In this paper, the Data Clustering algorithm based on Multi-objective Periodic Bacterial Foraging Optimization with two Learning Archives (DC-MPBFOLA) is proposed. Specifically, first, to reduce the high computing complexity of the original BFO, periodic BFO is employed as the basic algorithmic framework. Then transfer the periodic BFO into a multi-objective type. Second, two learning strategies are proposed based on the two learning archives to guide the bacterial swarm to move in a better direction. On the one hand, the global best is selected from the global learning archive according to the convergence index and diversity index. On the other hand, the personal best is selected from the personal learning archive according to the sum of weighted objectives. According to the aforementioned learning strategies, a chemotaxis operation is designed. Third, an elite learning strategy is designed to provide fresh power to the objects in two learning archives. When the objects in these two archives do not change for two consecutive times, randomly initializing one dimension of objects can prevent the proposed algorithm from falling into local optima. Fourth, to validate the performance of the proposed algorithm, DC-MPBFOLA is compared with four state-of-art evolutionary multi-objective optimization algorithms and one classical clustering algorithm on evaluation indexes of datasets. To further verify the effectiveness and feasibility of designed strategies in DC-MPBFOLA, variants of DC-MPBFOLA are also proposed. Experimental results demonstrate that DC-MPBFOLA outperforms its competitors regarding all evaluation indexes and clustering partitions. These results also indicate that the designed strategies positively influence the performance improvement of the original BFO.

Keywords: data clustering, multi-objective optimization, bacterial foraging optimization, learning archives

Procedia PDF Downloads 2
257 Towards a Conscious Design in AI by Overcoming Dark Patterns

Authors: Ayse Arslan


One of the important elements underpinning a conscious design is the degree of toxicity in communication. This study explores the mechanisms and strategies for identifying toxic content by avoiding dark patterns. Given the breadth of hate and harassment attacks, this study explores a threat model and taxonomy to assist in reasoning about strategies for detection, prevention, mitigation, and recovery. In addition to identifying some relevant techniques such as nudges, automatic detection, or human-ranking, the study suggests the use of major metrics such as the overhead and friction of solutions on platforms and users or balancing false positives (e.g., incorrectly penalizing legitimate users) against false negatives (e.g., users exposed to hate and harassment) to maintain a conscious design towards fairness.

Keywords: AI, ML, algorithms, policy, system design

Procedia PDF Downloads 6
256 A Study of Permission-Based Malware Detection Using Machine Learning

Authors: Ratun Rahman, Rafid Islam, Akin Ahmed, Kamrul Hasan, Hasan Mahmud


Malware is becoming more prevalent, and several threat categories have risen dramatically in recent years. This paper provides a bird's-eye view of the world of malware analysis. The efficiency of five different machine learning methods (Naive Bayes, K-Nearest Neighbor, Decision Tree, Random Forest, and TensorFlow Decision Forest) combined with features picked from the retrieval of Android permissions to categorize applications as harmful or benign is investigated in this study. The test set consists of 1,168 samples (among these android applications, 602 are malware and 566 are benign applications), each consisting of 948 features (permissions). Using the permission-based dataset, the machine learning algorithms then produce accuracy rates above 80%, except the Naive Bayes Algorithm with 65% accuracy. Of the considered algorithms TensorFlow Decision Forest performed the best with an accuracy of 90%.

Keywords: android malware detection, machine learning, malware, malware analysis

Procedia PDF Downloads 9
255 The Research of the Game Interface Improvement Due to the Game Operation Dilemma of Player in the Side-Scrolling Shooting Game

Authors: Shih-Chieh Liao, Cheng-Yan Shuai


The feature of a side-scrolling shooting game(2D STG )is facing the surrounding enemy and dodging the barrage on the entire screen while the player moves from the left to the right side of the game. In this kind of game, the player will be in trouble when they are trying to do the complicated operate because of the physical and system limitations of the joystick. Player’s operation needs to be flexible when playing this kind of games; furthermore, the joystick interface has a feature that one operates links two game mechanics. These make the current joystick cannot satisfy the available 2D STG. Long-term observation of this study find out three kinds of the dilemma when player using the joystick in the 2D STG. 1)Fail to shoot in the right place. 2) System misjudgment and operation error. 3)Over heavy joystick and system loading. Meanwhile, this study aggregates 11 types of joysticks in 3 classes by focusing on group interviewing. Then, we convene 5 different degree expert doing the learning assessment with SUS Score to screen out the best 3 types of each class, and then our study built up an experience environment for forging a new type of exclusive joystick for 2D STG. In the experimental stage, we convene the subject for a pre-test. First, using the survey to find out the novice in the subjects, then subjects play the ‘Metal Slug’ and record their efficiency and hit rate. When their score matches the normal distribution, we chose 120 subjects those are falls in the lower 25% to participate in our study. The experiment is simulating a dilemma that occur in the game, with subjects being separate to 3 treatment group and a control group and record their time efficiency and hit rate. The result shows that the exclusive joystick for 2D STG can improve the dilemma of novice, and the Type J-a-I is the most remarkable. Even though Type J-a-I has no distinction with the available joystick, it has better time efficiency to assist the player with the operation.

Keywords: arcade, joystick, 2D STG, usability

Procedia PDF Downloads 4
254 A Survey in Techniques for Imbalanced Intrusion Detection System Datasets

Authors: Najmeh Abedzadeh, Matthew Jacobs


An intrusion detection system (IDS) is a software application that monitors malicious activities and generates alerts if any are detected. However, most network activities in IDS datasets are normal, and the relatively few numbers of attacks make the available data imbalanced. Consequently, cyber-attacks can hide inside a large number of normal activities, and machine learning algorithms have difficulty learning and classifying the data correctly. In this paper, a comprehensive literature review is conducted on different types of algorithms for both implementing the IDS and methods in correcting the imbalanced IDS dataset. The most famous algorithms are machine learning (ML), deep learning (DL), synthetic minority over-sampling technique (SMOTE), and reinforcement learning (RL). Most of the research use the CSE-CIC-IDS2017, CSE-CIC-IDS2018, and NSL-KDD datasets for evaluating their algorithms.

Keywords: IDS, imbalanced datasets, sampling algorithms, big data

Procedia PDF Downloads 10
253 Blockchain-Resilient Framework for Cloud-Based Network Devices within the Architecture of Self-Driving Cars

Authors: Mirza Mujtaba Baig


Artificial Intelligence (AI) is evolving rapidly, and one of the areas in which this field has influenced is automation. The automobile, healthcare, education, and robotic industries deploy AI technologies constantly, and the automation of tasks is beneficial to allow time for knowledge-based tasks and also introduce convenience to everyday human endeavors. The paper reviews the challenges faced with the current implementations of autonomous self-driving cars by exploring the machine learning, robotics, and artificial intelligence techniques employed for the development of this innovation. The controversy surrounding the development and deployment of autonomous machines, e.g., vehicles, begs the need for the exploration of the configuration of the programming modules. This paper seeks to add to the body of knowledge of research assisting researchers in decreasing the inconsistencies in current programming modules. Blockchain is a technology of which applications are mostly found within the domains of financial, pharmaceutical, manufacturing, and artificial intelligence. The registering of events in a secured manner as well as applying external algorithms required for the data analytics are especially helpful for integrating, adapting, maintaining, and extending to new domains, especially predictive analytics applications.

Keywords: artificial intelligence, automation, big data, self-driving cars, machine learning, neural networking algorithm, blockchain, business intelligence

Procedia PDF Downloads 15
252 Exploratory Analysis of A Review of Nonexistence Polarity in Native Speech

Authors: Deawan Rakin Ahamed Remal, Sinthia Chowdhury, Sharun Akter Khushbu, Sheak Rashed Haider Noori


Native Speech to text synthesis has its own leverage for the purpose of mankind. The extensive nature of art to speaking different accents is common but the purpose of communication between two different accent types of people is quite difficult. This problem will be motivated by the extraction of the wrong perception of language meaning. Thus, many existing automatic speech recognition has been placed to detect text. Overall study of this paper mentions a review of NSTTR (Native Speech Text to Text Recognition) synthesis compared with Text to Text recognition. Review has exposed many text to text recognition systems that are at a very early stage to comply with the system by native speech recognition. Many discussions started about the progression of chatbots, linguistic theory another is rule based approach. In the Recent years Deep learning is an overwhelming chapter for text to text learning to detect language nature. To the best of our knowledge, In the sub continent a huge number of people speak in Bangla language but they have different accents in different regions therefore study has been elaborate contradictory discussion achievement of existing works and findings of future needs in Bangla language acoustic accent.

Keywords: TTR, NSTTR, text to text recognition, deep learning, natural language processing

Procedia PDF Downloads 18
251 A Conceptual Framework for Assessing the Development of Health Information Systems Enterprise Architecture Interoperability

Authors: Prosper Tafadzwa Denhere, Ephias Ruhode, Munyaradzi Zhou


Health Information Systems (HISs) interoperability is emerging to be the future of modern healthcare systems Enterprise Architecture (EA), where healthcare entities are seamlessly interconnected to share healthcare data. The reality that the healthcare industry has been characterised by an influx of fragmented stand-alone e-Health systems, which present challenges of healthcare information sharing across platforms, desires much attention for systems integration efforts. The lack of an EA conceptual framework resultantly crates the need for investigating an ideal solution to the objective of Health Information Systems interoperability development assessment. The study takes a qualitative exploratory approach through a design science research context. The research aims to study the various themes withdrawn from the literature that can help in the assessment of interoperable HISs development through a literature study. Themes derived from the study include HIS needs, HIS readiness, HIS constraints, and HIS technology integration elements and standards tied to the EA development architectural layers of The Open Group Architecture Framework (TOGAF) as an EA development methodology. Eventually, the themes were conceptualised into a framework reviewed by two experts. The essence of the study was to provide a framework within which interoperable EA of HISs should be developed.

Keywords: enterprise architecture, eHealth, health information systems, interoperability

Procedia PDF Downloads 17
250 Assessing the Effectiveness of Machine Learning Algorithms for Cyber Threat Intelligence Discovery from the Darknet

Authors: Azene Zenebe


Deep learning is a subset of machine learning which incorporates techniques for the construction of artificial neural networks and found to be useful for modeling complex problems with large dataset. Deep learning requires a very high power computational and longer time for training. By aggregating computing power, high performance computer (HPC) has emerged as an approach to resolving advanced problems and performing data-driven research activities. Cyber threat intelligence (CIT) is actionable information or insight an organization or individual uses to understand the threats that have, will, or are currently targeting the organization. Results of review of literature will be presented along with results of experimental study that compares the performance of tree-based and function-base machine learning including deep learning algorithms using secondary dataset collected from darknet.

Keywords: deep-learning, cyber security, cyber threat modeling, tree-based machine learning, function-based machine learning, data science

Procedia PDF Downloads 29
249 Parkinson’s Disease Detection Analysis through Machine Learning Approaches

Authors: Muhtasim Shafi Kader, Fizar Ahmed, Annesha Acharjee


Machine learning and data mining are crucial in health care, as well as medical information and detection. Machine learning approaches are now being utilized to improve awareness of a variety of critical health issues, including diabetes detection, neuron cell tumor diagnosis, COVID 19 identification, and so on. Parkinson’s disease is basically a disease for our senior citizens in Bangladesh. Parkinson's Disease indications often seem progressive and get worst with time. People got affected trouble walking and communicating with the condition advances. Patients can also have psychological and social vagaries, nap problems, hopelessness, reminiscence loss, and weariness. Parkinson's disease can happen in both men and women. Though men are affected by the illness at a proportion that is around partial of them are women. In this research, we have to get out the accurate ML algorithm to find out the disease with a predictable dataset and the model of the following machine learning classifiers. Therefore, nine ML classifiers are secondhand to portion study to use machine learning approaches like as follows, Naive Bayes, Adaptive Boosting, Bagging Classifier, Decision Tree Classifier, Random Forest classifier, XBG Classifier, K Nearest Neighbor Classifier, Support Vector Machine Classifier, and Gradient Boosting Classifier are used.

Keywords: naive bayes, adaptive boosting, bagging classifier, decision tree classifier, random forest classifier, XBG classifier, k nearest neighbor classifier, support vector classifier, gradient boosting classifier

Procedia PDF Downloads 36
248 Self-Supervised Learning for Hate-Speech Identification

Authors: Shrabani Ghosh


Automatic offensive language detection in social media has become a stirring task in today's NLP. Manual Offensive language detection is tedious and laborious work where automatic methods based on machine learning are only alternatives. Previous works have done sentiment analysis over social media in different ways such as supervised, semi-supervised, and unsupervised manner. Domain adaptation in a semi-supervised way has also been explored in NLP, where the source domain and the target domain are different. In domain adaptation, the source domain usually has a large amount of labeled data, while only a limited amount of labeled data is available in the target domain. Pretrained transformers like BERT, RoBERTa models are fine-tuned to perform text classification in an unsupervised manner to perform further pre-train masked language modeling (MLM) tasks. In previous work, hate speech detection has been explored in, which is a free speech platform described as a platform of extremist in varying degrees in online social media. In domain adaptation process, Twitter data is used as the source domain, and Gab data is used as the target domain. The performance of domain adaptation also depends on the cross-domain similarity. Different distance measure methods such as L2 distance, cosine distance, Maximum Mean Discrepancy (MMD), Fisher Linear Discriminant (FLD), and CORAL have been used to estimate domain similarity. Certainly, in-domain distances are small, and between-domain distances are expected to be large. The previous work finding shows that pretrain masked language model (MLM) fine-tuned with a mixture of posts of source and target domain gives higher accuracy. However, in-domain performance of the hate classifier on Twitter data accuracy is 71.78%, and out-of-domain performance of the hate classifier on Gab data goes down to 56.53%. Recently self-supervised learning got a lot of attention as it is more applicable when labeled data are scarce. Few works have already been explored to apply self-supervised learning on NLP tasks such as sentiment classification. Self-supervised language representation model ALBERTA focuses on modeling inter-sentence coherence and helps downstream tasks with multi-sentence inputs. Self-supervised attention learning approach shows better performance as it exploits extracted context word in the training process. In this work, a self-supervised attention mechanism has been proposed to detect hate speech on This framework initially classifies the Gab dataset in an attention-based self-supervised manner. On the next step, a semi-supervised classifier trained on the combination of labeled data from the first step and unlabeled data. The performance of the proposed framework will be compared with the results described earlier and also with optimized outcomes obtained from different optimization techniques.

Keywords: attention learning, language model, offensive language detection, self-supervised learning

Procedia PDF Downloads 24
247 Transfer Knowledge From Multiple Source Problems to a Target Problem in Genetic Algorithm

Authors: Terence Soule, Tami Al Ghamdi


To study how to transfer knowledge from multiple source problems to the target problem, we modeled the Transfer Learning (TL) process using Genetic Algorithms as the model solver. TL is the process that aims to transfer learned data from one problem to another problem. The TL process aims to help Machine Learning (ML) algorithms find a solution to the problems. The Genetic Algorithms (GA) give researchers access to information that we have about how the old problem is solved. In this paper, we have five different source problems, and we transfer the knowledge to the target problem. We studied different scenarios of the target problem. The results showed combined knowledge from multiple source problems improves the GA performance. Also, the process of combining knowledge from several problems results in promoting diversity of the transferred population.

Keywords: transfer learning, genetic algorithm, evolutionary computation, source and target

Procedia PDF Downloads 16
246 Partial Knowledge Transfer Between the Source Problem and the Target Problem in Genetic Algorithms

Authors: Terence Soule, Tami Al Ghamdi


To study how the partial knowledge transfer may affect the Genetic Algorithm (GA) performance, we model the Transfer Learning (TL) process using GA as the model solver. The objective of the TL is to transfer the knowledge from one problem to another related problem. This process imitates how humans think in their daily life. In this paper, we proposed to study a case where the knowledge transferred from the S problem has less information than what the T problem needs. We sampled the transferred population using different strategies of TL. The results showed transfer part of the knowledge is helpful and speeds the GA process of finding a solution to the problem.

Keywords: transfer learning, partial transfer, evolutionary computation, genetic algorithm

Procedia PDF Downloads 18
245 Intelligent Agent-Based Model for the 5G mmWave O2I Technology Adoption

Authors: Robert Joseph M. Licup


The deployment of the fifth-generation (5G) mobile system through mmWave frequencies is the new solution in the requirement to provide higher bandwidth readily available for all users. The usage pattern of the mobile users has moved towards either the work from home or online classes set-up because of the pandemic. Previous mobile technologies can no longer meet the high speed, and bandwidth requirement needed, given the drastic shift of transactions to the home. The millimeter-wave (mmWave) underutilized frequency is utilized by the fifth-generation (5G) cellular networks that support multi-gigabit-per-second (Gbps) transmission. However, due to its short wavelengths, high path loss, directivity, blockage sensitivity, and narrow beamwidth are some of the technical challenges that need to be addressed. Different tools, technologies, and scenarios are explored to support network design, accurate channel modeling, implementation, and deployment effectively. However, there is a big challenge on how the consumer will adopt this solution and maximize the benefits offered by the 5G Technology. This research proposes to study the intricacies of technology diffusion, individual attitude, behaviors, and how technology adoption will be attained. The agent based simulation model shaped by the actual applications, technology solution, and related literature was used to arrive at a computational model. The research examines the different attributes, factors, and intricacies that can affect each identified agent towards technology adoption.

Keywords: agent-based model, AnyLogic, 5G O21, 5G mmWave solutions, technology adoption

Procedia PDF Downloads 21
244 Using Wiki for Enhancing the Knowledge Transfer to Newcomers: An Experience Report

Authors: Hualter Oliveira Barbosa, Raquel Feitosa do Vale Cunha, Erika Muniz dos Santos, Fernanda Belmira Souza, Fabio Sousa, Luis Henrique Pascareli, Franciney de Oliveira Lima, Ana Cláudia Reis da Silva, Christiane Moreira de Almeida


Software development is intrinsic human-based knowledge-intensive. Due to globalization, software development has become a complex challenge and we usually face barriers related to knowledge management, team building, costly testing processes, especially in distributed settings. For this reason, several approaches have been proposed to minimize barriers caused by geographic distance. In this paper, we present as we use experimental studies to improve our knowledge management process using the Wiki system. According to the results, it was possible to identify learning preferences from our software projects leader team, organize and improve the learning experience of our Wiki and; facilitate collaboration by newcomers to improve Wiki with new contents available in the Wiki.

Keywords: mobile product, knowledge transfer, knowledge management process, wiki, GSD

Procedia PDF Downloads 35
243 Time Series Forecasting (TSF) Using Various Deep Learning Models

Authors: Jimeng Shi, Mahek Jain, Giri Narasimhan


Time Series Forecasting (TSF) is used to predict the target variables at a future time point based on the learning from previous time points. To keep the problem tractable, learning methods use data from a fixed-length window in the past as an explicit input. In this paper, we study how the performance of predictive models changes as a function of different look-back window sizes and different amounts of time to predict the future. We also consider the performance of the recent attention-based Transformer models, which have had good success in the image processing and natural language processing domains. In all, we compare four different deep learning methods (RNN, LSTM, GRU, and Transformer) along with a baseline method. The dataset (hourly) we used is the Beijing Air Quality Dataset from the UCI website, which includes a multivariate time series of many factors measured on an hourly basis for a period of 5 years (2010-14). For each model, we also report on the relationship between the performance and the look-back window sizes and the number of predicted time points into the future. Our experiments suggest that Transformer models have the best performance with the lowest Mean Average Errors (MAE = 14.599, 23.273) and Root Mean Square Errors (RSME = 23.573, 38.131) for most of our single-step and multi-steps predictions. The best size for the look-back window to predict 1 hour into the future appears to be one day, while 2 or 4 days perform the best to predict 3 hours into the future.

Keywords: air quality prediction, deep learning algorithms, time series forecasting, look-back window

Procedia PDF Downloads 31
242 Imbalanced Time-Series Data Regression Using Conditional Generative Adversarial Networks

Authors: Murtadha D. Hssayeni, Behnaz Ghoraani


During the collection of time-series data, many reasons lead to imbalanced and incomplete datasets. Consequently, when training deep convolutional models on these datasets, the models suffer from overfitting and lack generalizability to unseen data. In this paper, we investigated a new framework of Conditional Generative Adversarial Networks (cGANs) as a solution to improve the extrapolation and generalizability of the regression models in such datasets. We used an imbalanced synthetic dataset and two real-world datasets in Parkinson's disease (PD) application domain and Negative Affect (NA) estimation. In all scenarios, the developed cGAN demonstrated significantly better generalizability to unseen data samples than a traditional Convolutional Neural Network with an average improvement of 56% in mean absolute error in the case of the synthetic dataset, 34% in the PD dataset, and 18% in the NA dataset.

Keywords: regression, generative adversarial networks, imbalanced time-series data, incomplete data extrapolation

Procedia PDF Downloads 27
241 Virtualization and Visualization Based Driver Configuration in Operating System

Authors: Pavan Shah


In an Embedded system, Virtualization and visualization technology can provide us an effective response and measurable work in a software development environment. In addition to work of virtualization and virtualization can be easily deserved to provide the best resource sharing between real-time hardware applications and a healthy environment. However, the virtualization is noticeable work to minimize the I/O work and utilize virtualization & virtualization technology for either a software development environment (SDE) or a runtime environment of real-time embedded systems (RTMES) or real-time operating system (RTOS) eras. In this Paper, we particularly focus on virtualization and visualization overheads data of network which generates the I/O and implementation of standardized I/O (i.e., Virto), which can work as front-end network driver in a real-time operating system (RTOS) hardware module. Even there have been several work studies are available based on the virtualization operating system environment, but for the Virto on a general-purpose OS, my implementation is on the open-source Virto for a real-time operating system (RTOS). In this paper, the measurement results show that implementation which can improve the bandwidth and latency of memory management of the real-time operating system environment (RTMES) for getting more accuracy of the trained model.

Keywords: virtualization, visualization, network driver, operating system

Procedia PDF Downloads 51
240 Compact Dual-band 4-MIMO Antenna Elements for 5G Mobile Applications

Authors: Fayad Ghawbar


The significance of the Multiple Input Multiple Output (MIMO) system in the 5G wireless communication system is essential to enhance channel capacity and provide a high data rate resulting in a need for dual-polarization in vertical and horizontal. Furthermore, size reduction is critical in a MIMO system to deploy more antenna elements requiring a compact, low-profile design. A compact dual-band 4-MIMO antenna system has been presented in this paper with pattern and polarization diversity. The proposed single antenna structure has been designed using two antenna layers with a C shape in the front layer and a partial slot with a U-shaped cut in the ground to enhance isolation. The single antenna is printed on an FR4 dielectric substrate with an overall size of 18 mm×18 mm×1.6 mm. The 4-MIMO antenna elements were printed orthogonally on an FR4 substrate with a size dimension of 36 × 36 × 1.6 mm3 with zero edge-to-edge separation distance. The proposed compact 4-MIMO antenna elements resonate at 3.4-3.6 GHz and 4.8-5 GHz. The s-parameters measurement and simulation results agree, especially in the lower band with a slight frequency shift of the measurement results at the upper band due to fabrication imperfection. The proposed design shows isolation above -15 dB and -22 dB across the 4-MIMO elements. The MIMO diversity performance has been evaluated in terms of efficiency, ECC, DG, TARC, and CCL. The total and radiation efficiency were above 50 % across all parameters in both frequency bands. The ECC values were lower than 0.10, and the DG results were about 9.95 dB in all antenna elements. TARC results exhibited values lower than 0 dB with values lower than -25 dB in all MIMO elements at the dual-bands. Moreover, the channel capacity losses in the MIMO system were depicted using CCL with values lower than 0.4 Bits/s/Hz.

Keywords: compact antennas, MIMO antenna system, 5G communication, dual band, ECC, DG, TARC

Procedia PDF Downloads 38
239 Data Poisoning Attacks on Federated Learning and Preventive Measures

Authors: Beulah Rani Inbanathan


In the present era, it is vivid from the numerous outcomes that data privacy is being compromised in various ways. Machine learning is one technology that uses the centralized server, and then data is given as input which is being analyzed by the algorithms present on this mentioned server, and hence outputs are predicted. However, each time the data must be sent by the user as the algorithm will analyze the input data in order to predict the output, which is prone to threats. The solution to overcome this issue is federated learning, where the models alone get updated while the data resides on the local machine and does not get exchanged with the other local models. Nevertheless, even on these local models, there are chances of data poisoning, and it is crystal clear from various experiments done by many people. This paper delves into many ways where data poisoning occurs and the many methods through which it is prevalent that data poisoning still exists. It includes the poisoning attacks on IoT devices, Edge devices, Autoregressive model, and also, on Industrial IoT systems and also, few points on how these could be evadible in order to protect our data which is personal, or sensitive, or harmful when exposed.

Keywords: data poisoning, federated learning, Internet of Things, edge computing

Procedia PDF Downloads 6
238 Internal Audit Function Contributions to the External Audit

Authors: Douglas F. Prawitt, Nathan Y. Sharp, David A. Wood


Consistent with prior experimental and survey studies, we find that IAFs that spend more time directly assisting the external auditor is associated with lower external audit fees. Interestingly, we do not find evidence that external auditors reduce fees based on work previously performed by the IAF. We also find that the time spent assisting the external auditor has a greater negative effect on external audit fees than the time spent performing tasks upon which the auditor may rely but that are not performed as direct assistance to the external audit. Our results also show that previous proxies used to measure this relation is either not associated with or are negatively associated with our direct measures of how the IAF can contribute to the external audit and are highly positively associated with the size and the complexity of the organization. Thus, we conclude the disparate experimental and archival results may be attributable to issues surrounding the construct validity of measures used in previous archival studies and that when measures similar to those used in experimental studies are employed in archival tests, the archival results are consistent with experimental findings. Our research makes four primary contributions to the literature. First, we provide evidence that internal auditing contributes to a reduction in external audit fees. Second, we replicate and provide an explanation for why previous archival studies find that internal auditing has either no association with external audit fees or is associated with an increase in those fees: prior studies generally use proxies of internal audit contribution that do not adequately capture the intended construct. Third, our research expands on survey-based research (e.g., Oil Libya by separately examining the impact on the audit fee of the internal auditors’ work, indirectly assisting external auditors and internal auditors’ prior work upon which external auditors can rely. Finally, we extend prior research by using a new, independent data source to validate and extend prior studies. This data set also allows for a sample of examining the impact of internal auditing on the external audit fee and the use of a more comprehensive external audit fee model that better controls for determinants of the external audit fee.

Keywords: internal audit, contribution, external audit, function

Procedia PDF Downloads 43
237 In Agile Projects - Arithmetic Sequence is More Effective than Fibonacci Sequence to Use for Estimating the Implementation Effort of User Stories

Authors: Khaled Jaber


The estimation of effort in software development is a complex task. The traditional Waterfall approach used to develop software systems requires a lot of time to estimate the effort needed to implement user requirements. Agile manifesto, however, is currently more used in the industry than the Waterfall to develop software systems. In Agile, the user requirement is referred to as a user story. Agile teams mostly use the Fibonacci sequence 1, 2, 3, 5, 8, 11, etc. in estimating the effort needed to implement the user story. This work shows through analysis that the Arithmetic sequence, e.g., 3, 6, 9, 12, etc., is more effective than the Fibonacci sequence in estimating the user stories. This paper mathematically and visually proves the effectiveness of the Arithmetic sequence over the FB sequence.

Keywords: agie, scrum, estimation, fibonacci sequence

Procedia PDF Downloads 61
236 A Multi-Population DE with Adaptive Mutation and Local Search for Global Optimization

Authors: Zhoucheng Bao, Haiyan Zhu, Tingting Pang, Zuling Wang


This paper proposes a multi-population DE with adaptive mutation and local search for global optimization, named AMMADE. In order to better coordinate the cooperation between the populations and the rational use of resources. In AMMADE, the population is divided based on the Euclidean distance sorting method at each generation to appropriately coordinate the cooperation between subpopulations and the usage of resources, such that the best-performed subpopulation will get more computing resources in the next generation. Further, an adaptive local search strategy is employed on the best-performed subpopulation to achieve a balanced search. The proposed algorithm has been tested by solving optimization problems taken from CEC2014 benchmark problems. Experimental results show that our algorithm can achieve a competitive or better than related methods. The results also confirm the significance of devised strategies in the proposed algorithm.

Keywords: differential evolution, multi-mutation strategies, memetic algorithm, adaptive local search

Procedia PDF Downloads 65
235 Trusted Neural Network: Reversibility in Neural Networks for Network Integrity Verification

Authors: Malgorzata Schwab, Ashis Kumer Biswas


In this concept paper, we explore the topic of Reversibility in Neural Networks leveraged for Network Integrity Verification and crafted the term ''Trusted Neural Network'' (TNN), paired with the API abstraction around it, to embrace the idea formally. This newly proposed high-level generalizable TNN model builds upon the Invertible Neural Network architecture, trained simultaneously in both forward and reverse directions. This allows for the original system inputs to be compared with the ones reconstructed from the outputs in the reversed flow to assess the integrity of the end-to-end inference flow. The outcome of that assessment is captured as an Integrity Score. Concrete implementation reflecting the needs of specific problem domains can be derived from this general approach and is demonstrated in the experiments. The model aspires to become a useful practice in drafting high-level systems architectures which incorporate AI capabilities.

Keywords: trusted, neural, invertible, API

Procedia PDF Downloads 24
234 Developing API Economy: Associating Value to APIs and Microservices in an Enterprise

Authors: Mujahid Sultan


The IT industry has seen many transformations in the Software Development Life Cycle (SDLC) methodologies and development approaches. SDLCs range from waterfall to agile, and the development approaches from monolith to microservices. Management, orchestration, and monetization of microservices have created an API economy in the modern enterprise. There are two approaches to API design, code first and design first. Design first is gaining popularity in the industry as this allows capturing the API needs from the stakeholders rather than the development teams guesstimating the needs and associating a monetary value with the APIs and microservices. In this publication, we describe an approach to organizing and creating stakeholder needs and requirements for designing microservices and APIs.

Keywords: requirements engineering, enterprise architecture, APIs, microservices, DevOps, continuous delivery, continuous integration, stakeholder viewpoints

Procedia PDF Downloads 67
233 Multiple Fault Diagnosis in Digital Circuits using Critical Path Tracing and Enhanced Deduction Algorithm

Authors: Mohamed Mahmoud


This paper has developed an effect-cause analysis technique for fault diagnosis in digital circuits. The main algorithm of our technique is based on the Enhanced Deduction Algorithm, which processes the real response of the CUT to the applied test T to deduce the values of the internal lines. An experimental version of the algorithm has been implemented in C++. The code takes about 7592 lines. The internal values are determined based on the logic values under the permanent stuck-fault model. Using a backtracking strategy guarantees that the actual values are covered by at least one solution, or no solution is found.

Keywords: enhanced deduction algorithm, backtracking strategy, automatic test equipment, verfication

Procedia PDF Downloads 51
232 Quality Assessment of New Zealand Manuka Honeys Using Hyperspectral Imaging Combined with Deep Learning

Authors: Hien Thi Dieu Truong, Pullanagari Reddy, Marlon M. Reis, Richard Archer


New Zealand mānuka honey is a honeybee product derived mainly from Leptospermum scoparium nectar. The potent antibacterial activity of mānuka honey derives principally from methylglyoxal (MGO), in addition to the hydrogen peroxide and other lesser activities present in all honeys. MGO is formed from dihydroxyacetone (DHA) unique to L. scoparium nectar. Mānuka honey also has an idiosyncratic phenolic profile useful as a chemical maker. Authentic mānuka honey is highly valuable, but almost all honeys are formed from natural mixtures of nectars harvested by a hive over a time period. Once diluted by other nectars, mānuka honey irrevocably loses value. We aimed to apply hyperspectral imaging to honey frames before bulk extraction to minimisedilution of genuine mānuka by other honeys andensure authenticity at the source. This technology is non-destructive and suitable for an industrial setting. Chemometrics usingPartial Least Squares (PLS) and Support Vector Machine (SVM) showed limited efficacy in interpreting chemical footprints due to large non-linear relationships between predictor and predictand in a large sample set, likely due to honey quality variability across geographic regions. Therefore, an advanced modeling approach, one-dimensional convolutional neural networks (1D-CNN), was investigated for analysinghyperspectral data for extraction of biochemical information from honeys. The 1D-CNN model showed superior prediction of honey quality(R2 = 0.79, RMSE = 2.201) to PLS (R2 = 0.66, RMSE = 2.607) and SVM (R2 = 0.66, RMSE = 2.559). Classification of mono-floral manuka honeys from multi-floral and non-manuka honeys exceeded 90 % accuracy for all models tried. This study reveals the potential of 1D-CNN modeling for evaluating honey authenticity.

Keywords: authenticity, 1D-CNN, mānuka honey, quality

Procedia PDF Downloads 47
231 Requirement Engineering Within Open Source Software Development: A Case Study

Authors: Kars Beek, Remco Groeneveld, Sjaak Brinkkemper


Although there is much literature available on requirement documentation in traditional software development, few studies have been conducted about this topic in open source software development. While open-source software development is becoming more important, the software development processes are often not as structured as corporate software development processes. Papers show that communities, creating open-source software, often lack structure and documentation. However, most recent studies about this topic are often ten or more years old. Therefore, this research has been conducted to determine if the lack of structure and documentation in requirement engineering is currently still the situation in these communities. Three open-source products have been chosen as subjects for conducting this research. The data for this research was gathered based on interviews, observations, and analyses of feature proposals and issue tracking tools. In this paper, we present a comparison and an analysis of the different methods used for requirements documentation to understand the current practices of requirements documentation in open source software development.

Keywords: case study, open source software, open source software development, requirement elicitation, requirement engineering

Procedia PDF Downloads 9
230 Effects of Screen Time on Children from a Systems Engineering Perspective

Authors: Misagh Faezipour


This paper explores the effects of screen time on children from a systems engineering perspective. We reviewed literature from several related work on effects of screen time on children to explore all factors and interrelationships that would impact children that are subjected to using long screen times. Factors such as kids age, parent attitudes, parent screen time influence, amount of time kids spent with technology, psychosocial and physical health outcome, reduced mental imagery, problem solving and adaptive thinking skills, obesity, unhealthy diet, depressive symptoms, health problems, disruption in sleep behavior, decrease in physical activities, problematic relationship with mothers, language, social, emotional delays, are examples of some factors that could be either a cause or effect of screen time. A systems engineering perspective is used to explore all the factors and factor relationships that were discovered through literature. A causal model is used to illustrate a graphical representation of these factors and their relationships. Through the causal model, the factors with the highest impacts can be realized. The future work would be to develop a system dynamics model to view the dynamic behavior of the relationships and observe the impact of changes in different factors in the model. The different changes on the input of the model, such as a healthier diet or obesity rate, would depict the effect of the screen time in the model and portray the effect on the children’s health and other factors that are important, which also works as a decision support tool.

Keywords: screen time, systems engineering, causal model, system dynamics, children

Procedia PDF Downloads 71
229 Safety Validation of Black-Box Autonomous Systems: A Multi-Fidelity Reinforcement Learning Approach

Authors: Jared Beard, Ali Baheri


As autonomous systems become more prominent in society, ensuring their safe application becomes increasingly important. This is clearly demonstrated with autonomous cars traveling through a crowded city or robots traversing a warehouse with heavy equipment. Human environments can be complex, having high dimensional state and action spaces. This gives rise to two problems. One being that analytic solutions may not be possible. The other is that in simulation based approaches, searching the entirety of the problem space could be computationally intractable, ruling out formal methods. To overcome this, approximate solutions may seek to find failures or estimate their likelihood of occurrence. One such approach is adaptive stress testing (AST) which uses reinforcement learning to induce failures in the system. The premise of which is that a learned model can be used to help find new failure scenarios, making better use of simulations. In spite of these failures AST fails to find particularly sparse failures and can be inclined to find similar solutions to those found previously. To help overcome this, multi-fidelity learning can be used to alleviate this overuse of information. That is, information in lower fidelity can simulations can be used to build up samples less expensively, and more effectively cover the solution space to find a broader set of failures. Recent work in multi-fidelity learning has passed information bidirectionally using “knows what it knows” (KWIK) reinforcement learners to minimize the number of samples in high fidelity simulators (thereby reducing computation time and load). The contribution of this work, then, is development of the bidirectional multi-fidelity AST framework. Such an algorithm, uses multi-fidelity KWIK learners in an adversarial context to find failure modes. Thus far, a KWIK learner has been used to train an adversary in a grid world to prevent an agent from reaching its goal; thus demonstrating the utility of KWIK learners in an AST framework. The next step is implementation of the bidirectional multi-fidelity AST framework described. Testing will be conducted in a grid world containing an agent attempting to reach a goal position and adversary tasked with intercepting the agent as demonstrated previously. Fidelities will be modified by adjusting the size of a time-step, with higher-fidelity effectively allowing for more responsive closed loop feedback. Results will compare the single KWIK AST learner with the multi-fidelity algorithm with respect to number of samples, distinct failure modes found, and relative effect of learning after a number of trials.

Keywords: multi-fidelity reinforcement learning, multi-fidelity simulation, safety validation, falsification

Procedia PDF Downloads 20