COVID-19 is an epidemic disease that has threatened all the people at worldwide scale and eventua... more COVID-19 is an epidemic disease that has threatened all the people at worldwide scale and eventually became a pandemic It is a crucial task to differentiate COVID-19-affected patients from healthy patient populations. The need for technology enabled solutions is pertinent and this paper proposes a deep learning model for detection of COVID-19 using Chest X-Ray (CXR) images. In this research work, we provide insights on how to build robust deep learning based models for COVID-19 CXR image classification from Normal and Pneumonia affected CXR images. We contribute a methodical escort on preparation of data to produce a robust deep learning model. The paper prepared datasets by refactoring, using images from several datasets for ameliorate training of deep model. These recently published datasets enable us to build our own model and compare by using pretrained models. The proposed experiments show the ability to work effectively to classify COVID-19 patients utilizing CXR. The empirical work, which uses a 3 convolutional layer based Deep Neural Network called "DeepCOVNet" to classify CXR images into 3 classes: COVID-19, Normal and Pneumonia cases, yielded an accuracy of 96.77% and a F1-score of 0.96 on two different combination of datasets.
International Journal of Mathematical, Engineering and Management Sciences
Over the last decade, data sharing has become eminent in each aspect of the daily routine chores ... more Over the last decade, data sharing has become eminent in each aspect of the daily routine chores of industries and research alike. Traditionally, all data sharing platform depend on trusted third parties (TTP), owing to which they lack trust, security, immutability and transparency. But, with the advent of blockchain technology, the data sharing has got a whole new dimensionality. Blockchain is a distributed and decentralized ledger that records the source of a digital resources. The secure features of blockchain have helped it gain popularity and application in a variety of domains including sustainable manufacturing. It can aid in customer and product tracking, supply chain, quality checks, etc. Blockchain can further strengthen how products can be designed, engineered, manufactured, dispatched and tracked in the revolutionized Industry 4.0 initiative. All these activities involve sharing of voluminous data. Hence, this paper presents an efficient data-sharing system which takes t...
Electroencephalogram signals capture the brain electrical activity and provide factual cues to ex... more Electroencephalogram signals capture the brain electrical activity and provide factual cues to examine the current condition of a person which can be efficacious to understand and analyze the perfo...
Human action recognition refers to the classification of human action from video clips automatica... more Human action recognition refers to the classification of human action from video clips automatically. Images extracted from the video clips at regular time interval are processed to identify the action contained in them. This is done by comparing these images with images taken from appropriate standard action databases. Thus, human action recognition becomes the task of verifying the similarity between two images. This paper proposes mutual difference score as a measure of similarity between two images. The proposed measure has been validated using the Weizmann and KTH datasets.
Medical data contains sensitive information and can have big impact if stolen. The term medical i... more Medical data contains sensitive information and can have big impact if stolen. The term medical identity theft is coined for such thefts. Medical records are collected from various sources like hospitals, diagnostic labs, physicians, pharmacy and health insurance companies and includes all details of patient including his demographic information, test reports like X – rays, CT scans, MRIs, etc. With the advent of digitization, these medical records are now stored in digital form to make access and sharing easier. However, storing and sharing these data electronically opens the threat of data theft and misuse. Health insurance companies often bear the brunt of fraudulent claims based on stolen medical data. So, the current need is to enable storing and sharing these data with security and make a prohibition on making copies of such data. Hence, considering the importance of healthcare data, the use of blockchain technology can be promising to maintain the security, privacy, immutabil...
This paper introduces an approach to reinforcement learning by cooperating agents using a variati... more This paper introduces an approach to reinforcement learning by cooperating agents using a variation of the Q-learning method. Q-learning is a model free method i.e. in this method agent does not need to predict future condition. The framework provided by approximation space makes it possible to minimize the overestimation caused by approximated Q-values. Due to overestimation learning capability of the algorithm is not consistent. It is observed that under this condition the ability to take a particular action is decreased. Therefore, by using the Rough Q-learning method the performance of the algorithm increases. This is shown by comparing the average Q values for Q learning and Rough Q learning by means of plots.
Reinforcement learning has received much attention in the past decades. The three forms of reinfo... more Reinforcement learning has received much attention in the past decades. The three forms of reinforcement learning algorithms are Actor Critic learning, Q learning and Reinforcement Comparison. Q-learning is a form of model-free reinforcement learning with one drawback that is the overestimation (Rising Q) problem. To solve this problem Rough Sets approach is used. This has lead to the modification of the traditional Q learning algorithms to a new form of Q learning namely, Rough Q learning. Actor Critic Learning have a separate memory structure to explicitly represent the policy independent of the value function. Another form of reinforcement learning is Reinforcement Comparison. Using reinforcement comparison method (RC), a reference reward is equated with an average of previously received rewards. The Actor Critic and RC method is also made better by using the rough set approach. The results of the study are in form of various plots for all three forms of reinforcement learning, t...
With the rapid advancement in technology, digitization of documents is gaining popularity. For di... more With the rapid advancement in technology, digitization of documents is gaining popularity. For digitization, the printed or handwritten text needs to be converted to a computer-readable form. For this, the document has to go through line detection, character extraction, recognition and finally conversion to a computer-readable form. A variety of methods have been proposed for the same. The paper proposes a method for text extraction and recognition which is based on a data set called as a learn file which is a vector representation of the images in the data set. Recognition is achieved by using the difference ratio between the input image and the learn file. The paper also presents two applications of the proposed method: text extraction from printed document and automatic number plate recognition. After recognition, the identified characters are written on to a text file.
COVID-19 is an epidemic disease that has threatened all the people at worldwide scale and eventua... more COVID-19 is an epidemic disease that has threatened all the people at worldwide scale and eventually became a pandemic It is a crucial task to differentiate COVID-19-affected patients from healthy patient populations. The need for technology enabled solutions is pertinent and this paper proposes a deep learning model for detection of COVID-19 using Chest X-Ray (CXR) images. In this research work, we provide insights on how to build robust deep learning based models for COVID-19 CXR image classification from Normal and Pneumonia affected CXR images. We contribute a methodical escort on preparation of data to produce a robust deep learning model. The paper prepared datasets by refactoring, using images from several datasets for ameliorate training of deep model. These recently published datasets enable us to build our own model and compare by using pretrained models. The proposed experiments show the ability to work effectively to classify COVID-19 patients utilizing CXR. The empirical work, which uses a 3 convolutional layer based Deep Neural Network called "DeepCOVNet" to classify CXR images into 3 classes: COVID-19, Normal and Pneumonia cases, yielded an accuracy of 96.77% and a F1-score of 0.96 on two different combination of datasets.
International Journal of Mathematical, Engineering and Management Sciences
Over the last decade, data sharing has become eminent in each aspect of the daily routine chores ... more Over the last decade, data sharing has become eminent in each aspect of the daily routine chores of industries and research alike. Traditionally, all data sharing platform depend on trusted third parties (TTP), owing to which they lack trust, security, immutability and transparency. But, with the advent of blockchain technology, the data sharing has got a whole new dimensionality. Blockchain is a distributed and decentralized ledger that records the source of a digital resources. The secure features of blockchain have helped it gain popularity and application in a variety of domains including sustainable manufacturing. It can aid in customer and product tracking, supply chain, quality checks, etc. Blockchain can further strengthen how products can be designed, engineered, manufactured, dispatched and tracked in the revolutionized Industry 4.0 initiative. All these activities involve sharing of voluminous data. Hence, this paper presents an efficient data-sharing system which takes t...
Electroencephalogram signals capture the brain electrical activity and provide factual cues to ex... more Electroencephalogram signals capture the brain electrical activity and provide factual cues to examine the current condition of a person which can be efficacious to understand and analyze the perfo...
Human action recognition refers to the classification of human action from video clips automatica... more Human action recognition refers to the classification of human action from video clips automatically. Images extracted from the video clips at regular time interval are processed to identify the action contained in them. This is done by comparing these images with images taken from appropriate standard action databases. Thus, human action recognition becomes the task of verifying the similarity between two images. This paper proposes mutual difference score as a measure of similarity between two images. The proposed measure has been validated using the Weizmann and KTH datasets.
Medical data contains sensitive information and can have big impact if stolen. The term medical i... more Medical data contains sensitive information and can have big impact if stolen. The term medical identity theft is coined for such thefts. Medical records are collected from various sources like hospitals, diagnostic labs, physicians, pharmacy and health insurance companies and includes all details of patient including his demographic information, test reports like X – rays, CT scans, MRIs, etc. With the advent of digitization, these medical records are now stored in digital form to make access and sharing easier. However, storing and sharing these data electronically opens the threat of data theft and misuse. Health insurance companies often bear the brunt of fraudulent claims based on stolen medical data. So, the current need is to enable storing and sharing these data with security and make a prohibition on making copies of such data. Hence, considering the importance of healthcare data, the use of blockchain technology can be promising to maintain the security, privacy, immutabil...
This paper introduces an approach to reinforcement learning by cooperating agents using a variati... more This paper introduces an approach to reinforcement learning by cooperating agents using a variation of the Q-learning method. Q-learning is a model free method i.e. in this method agent does not need to predict future condition. The framework provided by approximation space makes it possible to minimize the overestimation caused by approximated Q-values. Due to overestimation learning capability of the algorithm is not consistent. It is observed that under this condition the ability to take a particular action is decreased. Therefore, by using the Rough Q-learning method the performance of the algorithm increases. This is shown by comparing the average Q values for Q learning and Rough Q learning by means of plots.
Reinforcement learning has received much attention in the past decades. The three forms of reinfo... more Reinforcement learning has received much attention in the past decades. The three forms of reinforcement learning algorithms are Actor Critic learning, Q learning and Reinforcement Comparison. Q-learning is a form of model-free reinforcement learning with one drawback that is the overestimation (Rising Q) problem. To solve this problem Rough Sets approach is used. This has lead to the modification of the traditional Q learning algorithms to a new form of Q learning namely, Rough Q learning. Actor Critic Learning have a separate memory structure to explicitly represent the policy independent of the value function. Another form of reinforcement learning is Reinforcement Comparison. Using reinforcement comparison method (RC), a reference reward is equated with an average of previously received rewards. The Actor Critic and RC method is also made better by using the rough set approach. The results of the study are in form of various plots for all three forms of reinforcement learning, t...
With the rapid advancement in technology, digitization of documents is gaining popularity. For di... more With the rapid advancement in technology, digitization of documents is gaining popularity. For digitization, the printed or handwritten text needs to be converted to a computer-readable form. For this, the document has to go through line detection, character extraction, recognition and finally conversion to a computer-readable form. A variety of methods have been proposed for the same. The paper proposes a method for text extraction and recognition which is based on a data set called as a learn file which is a vector representation of the images in the data set. Recognition is achieved by using the difference ratio between the input image and the learn file. The paper also presents two applications of the proposed method: text extraction from printed document and automatic number plate recognition. After recognition, the identified characters are written on to a text file.
Uploads
Papers by Shamama Anwar