John Ayeelyan - Academia.edu (original) (raw)
Papers by John Ayeelyan
Research Square (Research Square), Sep 27, 2022
Federated learning is a multiple device collaboration setup designed to solve machine learning pr... more Federated learning is a multiple device collaboration setup designed to solve machine learning problems under framework for aggregation and knowledge transfer in distributed local data. This distributed model ensures the privacy of data at each local node. Owing to its relevance, there has been extensive research activities and outcomes in federated learning with expanded applicability to different areas by the research community. As such, there is a vast research archive made available by the community with research work and articles related to the various aspects of federated learning such as applications, challenges, privacy, functionalities, and design. With respect to the function and design of federated learning, client selection, aggregation, knowledge transfer, management of distributed data (Non-IID), Incentive of data and communication cost are of paramount importance. Any effective design of federated learning requires these aspects to be well considered.There are numerous survey articles found among the available literature that focus on its application and challenges, opportunities, data privacy and protection, as well as on federated learning on internet of things, federated learning on edge computing, etc.
2023 15th International Conference on Knowledge and Smart Technology (KST)
2022 IEEE International Smart Cities Conference (ISC2), Sep 26, 2022
IEEE Access, 2022
In the traditional approach to a digital asset management system, the data processing mechanism i... more In the traditional approach to a digital asset management system, the data processing mechanism is not transparent or visible to the data owners since the data is managed solely by the service provider. With the rapid development of blockchain technology, the above issues can be resolved by leveraging the tamper-resistance and decentralization characteristics of blockchain. However, post the implementation of the EU General Data Protection Rules (GDPR) in 2018, the protection of data owners has taken center stage. This has led to several principles of personal data deletion, such as Storage Limit and the Right to Be Forgotten to conflict with the blockchain. It is also observed that, out of the various smart contracts deployed to manage digital assets, often only specific smart contracts are invoked, while the rest of the deployed smart contracts are rarely invoked, leading to smart contract designs exhibiting similar patterns with very little creativity. This current scenario has motivated us to propose SPChain, a smarter and private GDPR-compliant digital asset management framework enabled by blockchain. In this approach, a decentralized InterPlanetary File System has been adopted to solve the problem of SPOF. In addition, the combination of digital assets with artificial intelligence models has been proposed so as to make digital assets accessible to a larger number of applications and to enable better creativity. In this design, artificial intelligence models have been run in independent, virtualized containers and invoked through smart contracts. The proposed SPChain can be applied to the field of digital art management to provide a complete implementation based on the Hyperledger Fabric. Using this proposed framework, model developers, digital art creators, collectors, service providers, as well as third parties can not only benefit from securely managing digital assets and combining them with AI models, but also from simultaneously complying with the rights stipulated in the GDPR. During the course of the experiments conducted, the latency, throughput, and resource consumption of different functions in the smart contracts have been measured. After adjusting the batch timeout of the block and the maximum number of transactions in a block, the throughputs were observed to be about 500 TPS, with 10 to 15 TPS for reading and writing operations, respectively. The latency ranges were found to range from 0 to 7 seconds, with 2.5 to 5 seconds for reading and writing operations, respectively. INDEX TERMS Blockchain, AI smart contracts, GDPR, digital asset management.
Vehicles
With increasing urban population, there are more and more vehicles, causing traffic congestion. I... more With increasing urban population, there are more and more vehicles, causing traffic congestion. In order to solve this problem, the development of an efficient and fair intersection management system is an important issue. With the development of intelligent transportation systems, the computing efficiency of vehicles and vehicle-to-vehicle communications are becoming more advanced, which can be used to good advantage in developing smarter systems. As such, Autonomous Intersection Management (AIM) proposals have been widely discussed. This research proposes an intersection management system based on Advantage Actor-Critic (A2C) which is a type of reinforcement learning. This method can lead to a fair and efficient intersection resource allocation strategy being learned. In our proposed approach, we design a reward function and then use this reward function to encourage a fair allocation of intersection resources. The proposed approach uses a brake-safe control to ensure that autonom...
Algorithms
Deep learning and additive manufacturing have progressed together in the previous couple of decad... more Deep learning and additive manufacturing have progressed together in the previous couple of decades. Despite being one of the most promising technologies, they have several flaws that a collaborative effort may address. However, digital manufacturing has established itself in the current industrial revolution and it has slowed down quality control and inspection due to the different defects linked with it. Industry 4.0, the most recent industrial revolution, emphasizes the integration of intelligent production systems and current information technologies. As a result, deep learning has received a lot of attention and has been shown to be quite effective at understanding image data. This review aims to provide a cutting-edge deep learning application of the AM approach and application. This article also addresses the current issues of data privacy and security and potential solutions to provide a more significant dimension to future studies.
Super resolution (SR) image reconstruction is the process of combing several low resolution image... more Super resolution (SR) image reconstruction is the process of combing several low resolution images into a single high resolution image. The videos of the image change frame to frame. This paper is based on interpolation super-resolution method. An algorithm for enhancing the resolution of the scene through Segmentation of the video and cropping the required part of the scene, super-resolution using Interpolation, Regression, and Post-processing, is applied to the effective Super-resolution image output. Further object tracking and identification use the results of this work. We worked in traffic surveillance videos.
International Conference on Information Communication and Embedded Systems (ICICES2014), 2014
To evaluate near and far visual outcomes, subjective visual symptoms, and patient satisfaction wi... more To evaluate near and far visual outcomes, subjective visual symptoms, and patient satisfaction with AcrySof ® ReSTOR ® diffractive multifocal intraocular lenses (IOL), and to study the reasons for postoperative dissatisfaction. Methods: Twenty-three eyes of 19 patients received phacoemulsifications and implantation of AcrySof ® ReSTOR ® IOL. The main outcome measures, taken at postoperative 1 day, 1 week, 1 month, and 3 months, were uncorrected and corrected near and distant visual acuity, refractory errors, subjective visual symptoms (glare, halo, and night vision), and satisfaction. Results: At the 3-month postoperative visit, the mean uncorrected near and distant visual acuities were 0.59±0.24 (0.25±0.22 LogMAR unit) and 0.78±0.27 (0.13±0.10 LogMAR unit), respectively. In addition, patients' satisfaction with uncorrected near vision, intermediate vision, far vision, and general visual performance were better than their satisfaction with night vision. Glare and halos were reported as severe by only 10.2% and 5.3% of patients, respectively. The seven eyes with poor patient satisfaction included eyes with a high incidence of preoperative ocular diseases or preoperative and postoperative high corneal astigmatisms of more than 1.0 diopter. Conclusions: The AcrySof ® ReSTOR ® IOL demonstrated good near and distant visual acuity with good patient satisfaction. Previous ocular disease, corneal astigmatism less than 1.0 diopter, and patient lifestyle should be considered to enhance patient satisfaction.
International Journal of Computer Applications, 2013
In the past decade internet worked perfectly with distribution of the digital data for pictures, ... more In the past decade internet worked perfectly with distribution of the digital data for pictures, music and videos. Although digital data have many advantages over analogue data, the rightful ownership of the digital data source is at of risk. The copyright protection for digital media becomes an important issue of piracy. Watermarking is a very important field for copyrights of various electronic documents and multimedia. This paper presents a digital forensic watermarking method for authorization against copying or piracy of digital video. The core idea is to use biometric generated keys in the embedding process of watermark. The host video is first randomized by Heisenberg decomposition and Discrete Fourier Transform (DFT). The invisible watermark is embedded in the I-frame of the host video. The watermarks are embedded in the least significant bit (LSB) of the each block. The forensic watermark provides "The Chain of Custody" throughout the life cycle of the video distribution. The experimental result of test sequence demonstrates that the proposed work gives high security and robustness.
... the ACM, vol. 13, no. 7, pp. 422426, July 1970. [29] Harendra A.Alwis, Robin C.Doss, Praveen... more ... the ACM, vol. 13, no. 7, pp. 422426, July 1970. [29] Harendra A.Alwis, Robin C.Doss, Praveen S.Hewage , Morshed UU Chowdhury Topology Based Packet Marking for IP Traceback2006. © 2009 ACADEMY PUBLISHER
Recent Advances in Computer Science and Communications, 2021
Background: Predictive analytics has a multiplicity of statistical schemes from predictive modell... more Background: Predictive analytics has a multiplicity of statistical schemes from predictive modelling, data mining, machine learning. It scrutinizes present and chronological data to make predictions about expectations or if not unexplained measures. Most predictive models are used for business analytics to overcome loses and profit gaining. Predictive analytics is used to exploit the pattern in old and historical data. Objective: People used to follow some strategies for predicting stock value to invest in the more profit-gaining stocks and those strategies to search the stock market prices which are incorporated in some intelligent methods and tools. Such strategies will increase the investor’s profits and also minimize their risks. So prediction plays a vital role in stock market gaining and is also a very intricate and challenging process. Method: The proposed optimized strategies are the Deep Neural Network with Stochastic Gradient for stock prediction. The Neural Network is tra...
Blockchain and Machine Learning for e-Healthcare Systems
International Journal of Intelligent Engineering and Systems
Indexing and query processing is an emerging research field in spatio-temporal data. Most of the ... more Indexing and query processing is an emerging research field in spatio-temporal data. Most of the real-time applications such as location based services, fleet management, traffic prediction and radio frequency identification and sensor networks are based on spatio-temporal indexing and query processing. All the indexing and query processing applications is any one of the forms, such as spatio index access and supporting queries or spatio-temporal indexing method and support query or temporal dimension, while in spatial data it is considered as the second priority. In this paper, give the survey of the various uncertain indexing and query processing techniques. Most of the existing survey works on spatio-temporal are based on indexing methods and query processing, but presented separately. Both the indexing and querying are related, hence state-of-art of both the indexing and query processing techniques are considered together. This paper gives the details of spatio-temporal data classification, various types of indexing methods, query processing, application areas and research direction of spatio-temporal indexing and query processing.
In the past decade internet worked perfectly with distribution of the digital data for pictures, ... more In the past decade internet worked perfectly with distribution of the digital data for pictures, music and videos. Although digital data have many advantages over analogue data, the rightful ownership of the digital data source is at of risk. The copyright protection for digital media becomes an important issue of piracy. Watermarking is a very important field for copyrights of various electronic documents and multimedia. This paper presents a digital forensic watermarking method for authorization against copying or piracy of digital video. The core idea is to use biometric generated keys in the embedding process of watermark. The host video is first randomized by Heisenberg decomposition and Discrete Fourier Transform (DFT). The invisible watermark is embedded in the I-frame of the host video. The watermarks are embedded in the least significant bit (LSB) of the each block. The forensic watermark provides " The Chain of Custody " throughout the life cycle of the video distribution. The experimental result of test sequence demonstrates that the proposed work gives high security and robustness.
Abstract - The development of query processing on spatiotemporal network of objects moving on dyn... more Abstract - The development of query processing on spatiotemporal
network of objects moving on dynamic road
networks is a very challenging task because of a large
number of real environment applications dealing with
spatio-temporal objects with uncertain data. A moving
object update indexing technique, called Final Future
Trajectory Prediction Algorithm (FFTPA) is proposed,
which consists of an R-tree and moving object updations.
FFTPA not only supports queries related to the past
positions or trajectories of moving objects on uncertain
dynamic road networks, but also provides an efficient
update mechanism in dynamic road. The indexing technique
updates current positions and supports predictive query.
The performance analysis of this technique shows better
compared with FNR-Tree and PPFI trajectory query.
Abstract— The problem of identifying Distributed Denial of Service (DDoS) is one of the hardest t... more Abstract— The problem of identifying Distributed Denial of
Service (DDoS) is one of the hardest threats in the internet
security. It is important to protect the resource and trace
from the Denial of Service (DoS) attack, but it is difficult to
distinguish normal traffic and DoS attack traffic because the
DoS generally hide their identities/origins. Especially the
attackers often use incorrect or spoofed source IP address,
so tracing the source of the denial of service is hardest in
internet. Lot of techniques and methodologies are used to
trace the DDoS attacks. This paper presents some of the
mostly used predicting traceback techniques to solve the
problem. The main goal of this paper is appraise the
different traceback techniques of the DDoS. This paper
evaluates the different traceback methods of the
classification.
Research Square (Research Square), Sep 27, 2022
Federated learning is a multiple device collaboration setup designed to solve machine learning pr... more Federated learning is a multiple device collaboration setup designed to solve machine learning problems under framework for aggregation and knowledge transfer in distributed local data. This distributed model ensures the privacy of data at each local node. Owing to its relevance, there has been extensive research activities and outcomes in federated learning with expanded applicability to different areas by the research community. As such, there is a vast research archive made available by the community with research work and articles related to the various aspects of federated learning such as applications, challenges, privacy, functionalities, and design. With respect to the function and design of federated learning, client selection, aggregation, knowledge transfer, management of distributed data (Non-IID), Incentive of data and communication cost are of paramount importance. Any effective design of federated learning requires these aspects to be well considered.There are numerous survey articles found among the available literature that focus on its application and challenges, opportunities, data privacy and protection, as well as on federated learning on internet of things, federated learning on edge computing, etc.
2023 15th International Conference on Knowledge and Smart Technology (KST)
2022 IEEE International Smart Cities Conference (ISC2), Sep 26, 2022
IEEE Access, 2022
In the traditional approach to a digital asset management system, the data processing mechanism i... more In the traditional approach to a digital asset management system, the data processing mechanism is not transparent or visible to the data owners since the data is managed solely by the service provider. With the rapid development of blockchain technology, the above issues can be resolved by leveraging the tamper-resistance and decentralization characteristics of blockchain. However, post the implementation of the EU General Data Protection Rules (GDPR) in 2018, the protection of data owners has taken center stage. This has led to several principles of personal data deletion, such as Storage Limit and the Right to Be Forgotten to conflict with the blockchain. It is also observed that, out of the various smart contracts deployed to manage digital assets, often only specific smart contracts are invoked, while the rest of the deployed smart contracts are rarely invoked, leading to smart contract designs exhibiting similar patterns with very little creativity. This current scenario has motivated us to propose SPChain, a smarter and private GDPR-compliant digital asset management framework enabled by blockchain. In this approach, a decentralized InterPlanetary File System has been adopted to solve the problem of SPOF. In addition, the combination of digital assets with artificial intelligence models has been proposed so as to make digital assets accessible to a larger number of applications and to enable better creativity. In this design, artificial intelligence models have been run in independent, virtualized containers and invoked through smart contracts. The proposed SPChain can be applied to the field of digital art management to provide a complete implementation based on the Hyperledger Fabric. Using this proposed framework, model developers, digital art creators, collectors, service providers, as well as third parties can not only benefit from securely managing digital assets and combining them with AI models, but also from simultaneously complying with the rights stipulated in the GDPR. During the course of the experiments conducted, the latency, throughput, and resource consumption of different functions in the smart contracts have been measured. After adjusting the batch timeout of the block and the maximum number of transactions in a block, the throughputs were observed to be about 500 TPS, with 10 to 15 TPS for reading and writing operations, respectively. The latency ranges were found to range from 0 to 7 seconds, with 2.5 to 5 seconds for reading and writing operations, respectively. INDEX TERMS Blockchain, AI smart contracts, GDPR, digital asset management.
Vehicles
With increasing urban population, there are more and more vehicles, causing traffic congestion. I... more With increasing urban population, there are more and more vehicles, causing traffic congestion. In order to solve this problem, the development of an efficient and fair intersection management system is an important issue. With the development of intelligent transportation systems, the computing efficiency of vehicles and vehicle-to-vehicle communications are becoming more advanced, which can be used to good advantage in developing smarter systems. As such, Autonomous Intersection Management (AIM) proposals have been widely discussed. This research proposes an intersection management system based on Advantage Actor-Critic (A2C) which is a type of reinforcement learning. This method can lead to a fair and efficient intersection resource allocation strategy being learned. In our proposed approach, we design a reward function and then use this reward function to encourage a fair allocation of intersection resources. The proposed approach uses a brake-safe control to ensure that autonom...
Algorithms
Deep learning and additive manufacturing have progressed together in the previous couple of decad... more Deep learning and additive manufacturing have progressed together in the previous couple of decades. Despite being one of the most promising technologies, they have several flaws that a collaborative effort may address. However, digital manufacturing has established itself in the current industrial revolution and it has slowed down quality control and inspection due to the different defects linked with it. Industry 4.0, the most recent industrial revolution, emphasizes the integration of intelligent production systems and current information technologies. As a result, deep learning has received a lot of attention and has been shown to be quite effective at understanding image data. This review aims to provide a cutting-edge deep learning application of the AM approach and application. This article also addresses the current issues of data privacy and security and potential solutions to provide a more significant dimension to future studies.
Super resolution (SR) image reconstruction is the process of combing several low resolution image... more Super resolution (SR) image reconstruction is the process of combing several low resolution images into a single high resolution image. The videos of the image change frame to frame. This paper is based on interpolation super-resolution method. An algorithm for enhancing the resolution of the scene through Segmentation of the video and cropping the required part of the scene, super-resolution using Interpolation, Regression, and Post-processing, is applied to the effective Super-resolution image output. Further object tracking and identification use the results of this work. We worked in traffic surveillance videos.
International Conference on Information Communication and Embedded Systems (ICICES2014), 2014
To evaluate near and far visual outcomes, subjective visual symptoms, and patient satisfaction wi... more To evaluate near and far visual outcomes, subjective visual symptoms, and patient satisfaction with AcrySof ® ReSTOR ® diffractive multifocal intraocular lenses (IOL), and to study the reasons for postoperative dissatisfaction. Methods: Twenty-three eyes of 19 patients received phacoemulsifications and implantation of AcrySof ® ReSTOR ® IOL. The main outcome measures, taken at postoperative 1 day, 1 week, 1 month, and 3 months, were uncorrected and corrected near and distant visual acuity, refractory errors, subjective visual symptoms (glare, halo, and night vision), and satisfaction. Results: At the 3-month postoperative visit, the mean uncorrected near and distant visual acuities were 0.59±0.24 (0.25±0.22 LogMAR unit) and 0.78±0.27 (0.13±0.10 LogMAR unit), respectively. In addition, patients' satisfaction with uncorrected near vision, intermediate vision, far vision, and general visual performance were better than their satisfaction with night vision. Glare and halos were reported as severe by only 10.2% and 5.3% of patients, respectively. The seven eyes with poor patient satisfaction included eyes with a high incidence of preoperative ocular diseases or preoperative and postoperative high corneal astigmatisms of more than 1.0 diopter. Conclusions: The AcrySof ® ReSTOR ® IOL demonstrated good near and distant visual acuity with good patient satisfaction. Previous ocular disease, corneal astigmatism less than 1.0 diopter, and patient lifestyle should be considered to enhance patient satisfaction.
International Journal of Computer Applications, 2013
In the past decade internet worked perfectly with distribution of the digital data for pictures, ... more In the past decade internet worked perfectly with distribution of the digital data for pictures, music and videos. Although digital data have many advantages over analogue data, the rightful ownership of the digital data source is at of risk. The copyright protection for digital media becomes an important issue of piracy. Watermarking is a very important field for copyrights of various electronic documents and multimedia. This paper presents a digital forensic watermarking method for authorization against copying or piracy of digital video. The core idea is to use biometric generated keys in the embedding process of watermark. The host video is first randomized by Heisenberg decomposition and Discrete Fourier Transform (DFT). The invisible watermark is embedded in the I-frame of the host video. The watermarks are embedded in the least significant bit (LSB) of the each block. The forensic watermark provides "The Chain of Custody" throughout the life cycle of the video distribution. The experimental result of test sequence demonstrates that the proposed work gives high security and robustness.
... the ACM, vol. 13, no. 7, pp. 422426, July 1970. [29] Harendra A.Alwis, Robin C.Doss, Praveen... more ... the ACM, vol. 13, no. 7, pp. 422426, July 1970. [29] Harendra A.Alwis, Robin C.Doss, Praveen S.Hewage , Morshed UU Chowdhury Topology Based Packet Marking for IP Traceback2006. © 2009 ACADEMY PUBLISHER
Recent Advances in Computer Science and Communications, 2021
Background: Predictive analytics has a multiplicity of statistical schemes from predictive modell... more Background: Predictive analytics has a multiplicity of statistical schemes from predictive modelling, data mining, machine learning. It scrutinizes present and chronological data to make predictions about expectations or if not unexplained measures. Most predictive models are used for business analytics to overcome loses and profit gaining. Predictive analytics is used to exploit the pattern in old and historical data. Objective: People used to follow some strategies for predicting stock value to invest in the more profit-gaining stocks and those strategies to search the stock market prices which are incorporated in some intelligent methods and tools. Such strategies will increase the investor’s profits and also minimize their risks. So prediction plays a vital role in stock market gaining and is also a very intricate and challenging process. Method: The proposed optimized strategies are the Deep Neural Network with Stochastic Gradient for stock prediction. The Neural Network is tra...
Blockchain and Machine Learning for e-Healthcare Systems
International Journal of Intelligent Engineering and Systems
Indexing and query processing is an emerging research field in spatio-temporal data. Most of the ... more Indexing and query processing is an emerging research field in spatio-temporal data. Most of the real-time applications such as location based services, fleet management, traffic prediction and radio frequency identification and sensor networks are based on spatio-temporal indexing and query processing. All the indexing and query processing applications is any one of the forms, such as spatio index access and supporting queries or spatio-temporal indexing method and support query or temporal dimension, while in spatial data it is considered as the second priority. In this paper, give the survey of the various uncertain indexing and query processing techniques. Most of the existing survey works on spatio-temporal are based on indexing methods and query processing, but presented separately. Both the indexing and querying are related, hence state-of-art of both the indexing and query processing techniques are considered together. This paper gives the details of spatio-temporal data classification, various types of indexing methods, query processing, application areas and research direction of spatio-temporal indexing and query processing.
In the past decade internet worked perfectly with distribution of the digital data for pictures, ... more In the past decade internet worked perfectly with distribution of the digital data for pictures, music and videos. Although digital data have many advantages over analogue data, the rightful ownership of the digital data source is at of risk. The copyright protection for digital media becomes an important issue of piracy. Watermarking is a very important field for copyrights of various electronic documents and multimedia. This paper presents a digital forensic watermarking method for authorization against copying or piracy of digital video. The core idea is to use biometric generated keys in the embedding process of watermark. The host video is first randomized by Heisenberg decomposition and Discrete Fourier Transform (DFT). The invisible watermark is embedded in the I-frame of the host video. The watermarks are embedded in the least significant bit (LSB) of the each block. The forensic watermark provides " The Chain of Custody " throughout the life cycle of the video distribution. The experimental result of test sequence demonstrates that the proposed work gives high security and robustness.
Abstract - The development of query processing on spatiotemporal network of objects moving on dyn... more Abstract - The development of query processing on spatiotemporal
network of objects moving on dynamic road
networks is a very challenging task because of a large
number of real environment applications dealing with
spatio-temporal objects with uncertain data. A moving
object update indexing technique, called Final Future
Trajectory Prediction Algorithm (FFTPA) is proposed,
which consists of an R-tree and moving object updations.
FFTPA not only supports queries related to the past
positions or trajectories of moving objects on uncertain
dynamic road networks, but also provides an efficient
update mechanism in dynamic road. The indexing technique
updates current positions and supports predictive query.
The performance analysis of this technique shows better
compared with FNR-Tree and PPFI trajectory query.
Abstract— The problem of identifying Distributed Denial of Service (DDoS) is one of the hardest t... more Abstract— The problem of identifying Distributed Denial of
Service (DDoS) is one of the hardest threats in the internet
security. It is important to protect the resource and trace
from the Denial of Service (DoS) attack, but it is difficult to
distinguish normal traffic and DoS attack traffic because the
DoS generally hide their identities/origins. Especially the
attackers often use incorrect or spoofed source IP address,
so tracing the source of the denial of service is hardest in
internet. Lot of techniques and methodologies are used to
trace the DDoS attacks. This paper presents some of the
mostly used predicting traceback techniques to solve the
problem. The main goal of this paper is appraise the
different traceback techniques of the DDoS. This paper
evaluates the different traceback methods of the
classification.