2021 International Conference on COMmunication Systems & NETworkS (COMSNETS)
Various quality assessment parameters for multimedia traffic in the wireless network depends on r... more Various quality assessment parameters for multimedia traffic in the wireless network depends on reckoning Quality of Experience (QoE) from Quality of Service (QoS). Mean Opinion Score (MOS) is the extensively used network quality metric for integrated (data and video) traffic management and resource allocation. This work mainly studies an uplink underlay Dynamic Spectrum Access (DSA) optimization problem that utilizes the Deep Reinforcement Learning (DRL) algorithm for simultaneous QoE enhancement and interference management within a tolerable limit. A Resource Allocation Deep Deterministic Policy Gradient (RADDPG) algorithm is proposed for joint quality improvement and distortion maintenance. In this work, the Deterministic Policy Gradient method merges Deep Q Network (DQN) along with the policy gradient actor-critic framework to choose suitable actions for improving the learning process speed, stability and computation time therefore accomplishing precise estimations. Simulation result shows that the proposed RADDPG method outperforms the existing Q and DQN learning algorithm.
Antibiotic use for cardiovascular implantable devices (CIED) prophylaxis is well-accepted despite... more Antibiotic use for cardiovascular implantable devices (CIED) prophylaxis is well-accepted despite a paucity of data. Pre-procedural prophylaxis lowers the rate of CIED infections; however, data is lacking for intra- or post-procedural antibiotic use. Antibiotic-eluting envelopes (ENVELOPE) [TYRX®TM] have been shown to reduce post-procedural infections. Understanding implanter practices may provide insight as to the need for antibiotic stewardship. The purpose of this survey was to assess the practices of implanters nationally. A survey was completed by 150 implanters across the US. Participants were board certificated, implanters of CIEDs, with varying experience (1–25 years), in various hospital settings. Of the respondents, 97% reported routine use of systemic antibiotics pre-operatively. About two-thirds of implanters continue systemic antibiotics post-operatively, with half continuing antibiotics for >24 h; 83% of implanters add antibiotic to saline for the purpose of irrigat...
Proceedings of the 50th Annual IEEE/ACM International Symposium on Microarchitecture, 2017
Non-volatile memories (NVMs) have attracted signicant interest recently due to their high-density... more Non-volatile memories (NVMs) have attracted signicant interest recently due to their high-density, low static power, and persistence. There are, however, several challenges associated with building practical systems from NVMs, including limited write endurance and long latencies. Researchers have proposed a variety of architectural techniques which can achieve dierent tradeos between lifetime, performance and energy eciency; however, no individual technique can satisfy requirements for all applications and dierent objectives. Hence, we propose Memory Cocktail Therapy (MCT), a general, learning-based framework that adaptively chooses the best techniques for the current application and objectives. Specically, MCT performs four procedures to adapt the techniques to various scenarios. First, MCT formulates a high-dimensional conguration space from all dierent combinations of techniques. Second, MCT selects primary features from the conguration space with lasso regularization. Third, MCT estimates lifetime, performance and energy consumption using lightweight online predictors (eg. quadratic regression and gradient boosting) and a small set of congurations guided by the selected features. Finally, given the estimation of all congurations, MCT selects the optimal conguration based on the user-dened objectives. As a proof of concept, we test MCT's ability to guarantee dierent lifetime targets and achieve 95% of maximum performance, while minimizing energy consumption. We nd that MCT improves performance by 9.24% and reduces energy by 7.95% compared to the best static conguration. Moreover, the performance of MCT is 94.49% of the ideal conguration with only 5.3% more energy consumption.
This thesis is about using statistical methods for performance and power estimation which would a... more This thesis is about using statistical methods for performance and power estimation which would allow us to develop better scheduling algorithms and also more energy efficient systems. In many deployments, computer systems are underutilized – meaning that applications have performance requirements that demand less than full system capacity. Ideally, we would take advantage of this under-utilization by allocating system resources so that the performance requirements are met and energy is minimized. This optimization problem is complicated by the fact that the performance and power consumption of various system configurations are often application – or even input – dependent. Thus, practically, minimizing energy for a performance constraint requires fast, accurate estimations of application-dependent performance and power tradeoffs. We propose a set of algorithms for different scenarios to tackle this problem. First, we propose LEO, a probabilistic graphical model-based learning syste...
Participation of less-represented groups in clinical research has recently gained a great deal of... more Participation of less-represented groups in clinical research has recently gained a great deal of interest from organizations, as they have begun to recognize the importance of ensuring the safety of treatments and medications on a variety of people prior to use with the general population. One such group with the lowest amount of participation in clinical studies is minority women, as there are fewer minorities than non-minorities participating, and less women than men. One reason that has been suggested for why this occurs is that minorities are less trusting of doctors and the healthcare system. The purpose of this study in general was to determine the factors encouraging or discouraging women from clinical research, but the focus of this is on whether distrust is the root of the problem. This was accomplished by surveying women about this and other factors that may motivate their decisions to partake in clinical studies at a variety of locations, and performing statistical analy...
2017 IEEE International Conference on Autonomic Computing (ICAC), 2017
Independent applications co-scheduled on the same hardware will interfere with one another, affec... more Independent applications co-scheduled on the same hardware will interfere with one another, affecting performance in complicated ways. Predicting this interference is key to efficiently scheduling applications on shared hardware, but forming accurate predictions is difficult because there are many shared hardware features that could lead to the interference. In this paper we investigate machine learning approaches (specifically, regularization) to understand the relation between those hardware features and application interference. We propose ESP, a highly accurate and fast regularization technique for application interference prediction. To demonstrate this practicality, we implement ESP and integrate it into a scheduler for both single and multi-node Linux/x86 systems and compare the scheduling performance to state-of-the-art heuristics. We find that ESP-based schedulers increase throughput by 1.25-1.8× depending on the scheduling scenario. Additionally, we find that ESP's accurate predictions allow schedulers to avoid catastrophic decisions, which heuristic approaches fundamentally cannot detect.
New Insights into the Future of Pharmacoepidemiology and Drug Safety, 2021
The pharmacovigilance has been evolved as a professional and ethical practice in ensuring the saf... more The pharmacovigilance has been evolved as a professional and ethical practice in ensuring the safety of medicines. The Adverse Drug Reactions (ADRs) associated with the use of medicines including Anti-Tuberculous Therapy (ATT) through a robust system of pharmacovigilance helps in promoting the safety of patients at large. The occurrence of ADRs associated with the use of ATT is expected, a large number of medicines are combined and used for prolonged duration. The suspected ADRs associated with first line ATT are well documented. However, the drugs used in second line or multidrug resistant to tuberculosis (TB), namely bedaquiline, reported to cause QT prolongation in electrocardiogram reading as one of the most common ADRs. Therefore, early identification and prevention of ADRs during ATT is essential for promoting the rational use and reduce the burden of anti-microbial resistance, besides achieving better treatment outcomes.
Proceedings of the 46th International Symposium on Computer Architecture, 2019
Machine learning and artiicial intelligence are invaluable for computer systems optimization: as ... more Machine learning and artiicial intelligence are invaluable for computer systems optimization: as computer systems expose more resources for management, ML/AI is necessary for modeling these resources' complex interactions. The standard way to incorporate ML/AI into a computer system is to irst train a learner to accurately predict the system's behavior as a function of resource usageÐe.g., to predict energy eiciency as a function of core usageÐand then deploy the learned model as part of a systemÐe.g., a scheduler. In this paper, we show that (1) continued improvement of learning accuracy may not improve the systems result, but (2) incorporating knowledge of the systems problem into the learning process improves the systems results even though it may not improve overall accuracy. Speciically, we learn application performance and power as a function of resource usage with the systems goal of meeting latency constraints with minimal energy. We propose a novel generative model which improves learning accuracy given scarce data, and we propose a multi-phase sampling technique, which incorporates knowledge of the systems problem. Our results are both positive and negative. The generative model improves accuracy, even for state-of-the-art learning systems, but negatively impacts energy. Multi-phase sampling reduces energy consumption compared to the state-of-the-art, but does not improve accuracy. These results imply that learning for systems optimization may have reached a point of diminishing returns where accuracy improvements have little efect on the systems outcome. Thus we advocate that future work on learning for systems should de-emphasize accuracy and instead incorporate the system problem's structure into the learner. CCS CONCEPTS • Computing methodologies → Machine learning; • Computer systems organization → Heterogeneous (hybrid) systems; Embedded systems; Real-time system architecture; • Hardware → Chip-level power issues.
The binding free energy calculation of protein–ligand complexes is necessary for research into vi... more The binding free energy calculation of protein–ligand complexes is necessary for research into virus–host interactions and the relevant applications in drug discovery. However, many current computational methods of such calculations are either inefficient or inaccurate in practice. Utilizing implicit solvent models in the molecular mechanics generalized Born surface area (MM/GBSA) framework allows for efficient calculations without significant loss of accuracy. Here, GBNSR6, a new flavor of the generalized Born model, is employed in the MM/GBSA framework for measuring the binding affinity between SARS-CoV-2 spike protein and the human ACE2 receptor. A computational protocol is developed based on the widely studied Ras–Raf complex, which has similar binding free energy to SARS-CoV-2/ACE2. Two options for representing the dielectric boundary of the complexes are evaluated: one based on the standard Bondi radii and the other based on a newly developed set of atomic radii (OPT1), optimi...
New Insights into the Future of Pharmacoepidemiology and Drug Safety [Working Title], 2021
Medical devices are being used in healthcare facilities for diagnosis, monitoring, prevention and... more Medical devices are being used in healthcare facilities for diagnosis, monitoring, prevention and treatment of an array of diseases. To ensure user/patient safety associated with the medical devices being used in healthcare industry, it is of utmost importance to closely monitor the adverse events associated with the medical devices through a robust, sustainable and scaled surveillance. Materiovigilance Programme of India (MvPI) provides a reliable system to report adverse events associated with medical devices. Under MvPI, various modalities to report adverse events associated with medical devices have been developed. These modalities include an editable medical device adverse event reporting form, a toll-free helpline number and a field safety corrective action form (FSCA). FSCA form is used to notify the regulatory authority and healthcare professionals on corrective actions or recall by the manufacturer. Due to the emergence of the Coronavirus disease 2019 (COVID-19) pandemic, o...
Cognitive radio (CR) is a flexible wireless network that can solve the scarcity and underutilizat... more Cognitive radio (CR) is a flexible wireless network that can solve the scarcity and underutilization problem of the spectrum by permitting unlicensed users to access licensed bands. The dynamic nature of CR makes it more vulnerable in terms of security. This paper's emphasis is on the primary user emulation attack, which poses a severe threat to the spectrum sensing operation of CR. In this attack, a malicious user imitates the signal characteristics of a licensed user (primary user) to disguise its true identity. Although many survey papers enhance our knowledge of cognitive radio security, this paper is an attempt to culminate new findings with the old ones to keep up the pace of the research community. Finally, the paper summarizes with some recommendations and future strategies pertinent to energy-efficient and flexible security methods for next-generation wireless systems.
Background: The relationship between histological findings of preimplantation kidney biopsies and... more Background: The relationship between histological findings of preimplantation kidney biopsies and pump resistive index (RI) values during machine preservation and allograft survival remains unclear due to conflicting results between different studies. Objective: To determine the association between histological findings of preimplantation kidney biopsies and pump resistive index values and short-term graft survival (up to 3 years post-transplant). Methods: We performed a retrospective review of all patients who received a kidney-only transplant between January 1, 2009, and December 31, 2014, at our center and followed-up until December 1, 2016. Patients who died with a functioning allograft were excluded from analysis. Obsolete glomerulosclerosis was divided into 3 groups (< 10%, 10-20%, and > 20%), intimal arterial narrowing was divided into 3 groups (< 10%, 11-25% and > 25%) and RI was divided into 5 groups (0.00-0.09, 0.10-0.19, 0.20-0.29, 0.30-0.39 and 0.40=0.49). The log-rank test was used to test if there is a significant difference in different RI groups in terms of the probability of survival.
2021 International Conference on COMmunication Systems & NETworkS (COMSNETS)
Various quality assessment parameters for multimedia traffic in the wireless network depends on r... more Various quality assessment parameters for multimedia traffic in the wireless network depends on reckoning Quality of Experience (QoE) from Quality of Service (QoS). Mean Opinion Score (MOS) is the extensively used network quality metric for integrated (data and video) traffic management and resource allocation. This work mainly studies an uplink underlay Dynamic Spectrum Access (DSA) optimization problem that utilizes the Deep Reinforcement Learning (DRL) algorithm for simultaneous QoE enhancement and interference management within a tolerable limit. A Resource Allocation Deep Deterministic Policy Gradient (RADDPG) algorithm is proposed for joint quality improvement and distortion maintenance. In this work, the Deterministic Policy Gradient method merges Deep Q Network (DQN) along with the policy gradient actor-critic framework to choose suitable actions for improving the learning process speed, stability and computation time therefore accomplishing precise estimations. Simulation result shows that the proposed RADDPG method outperforms the existing Q and DQN learning algorithm.
Antibiotic use for cardiovascular implantable devices (CIED) prophylaxis is well-accepted despite... more Antibiotic use for cardiovascular implantable devices (CIED) prophylaxis is well-accepted despite a paucity of data. Pre-procedural prophylaxis lowers the rate of CIED infections; however, data is lacking for intra- or post-procedural antibiotic use. Antibiotic-eluting envelopes (ENVELOPE) [TYRX®TM] have been shown to reduce post-procedural infections. Understanding implanter practices may provide insight as to the need for antibiotic stewardship. The purpose of this survey was to assess the practices of implanters nationally. A survey was completed by 150 implanters across the US. Participants were board certificated, implanters of CIEDs, with varying experience (1–25 years), in various hospital settings. Of the respondents, 97% reported routine use of systemic antibiotics pre-operatively. About two-thirds of implanters continue systemic antibiotics post-operatively, with half continuing antibiotics for >24 h; 83% of implanters add antibiotic to saline for the purpose of irrigat...
Proceedings of the 50th Annual IEEE/ACM International Symposium on Microarchitecture, 2017
Non-volatile memories (NVMs) have attracted signicant interest recently due to their high-density... more Non-volatile memories (NVMs) have attracted signicant interest recently due to their high-density, low static power, and persistence. There are, however, several challenges associated with building practical systems from NVMs, including limited write endurance and long latencies. Researchers have proposed a variety of architectural techniques which can achieve dierent tradeos between lifetime, performance and energy eciency; however, no individual technique can satisfy requirements for all applications and dierent objectives. Hence, we propose Memory Cocktail Therapy (MCT), a general, learning-based framework that adaptively chooses the best techniques for the current application and objectives. Specically, MCT performs four procedures to adapt the techniques to various scenarios. First, MCT formulates a high-dimensional conguration space from all dierent combinations of techniques. Second, MCT selects primary features from the conguration space with lasso regularization. Third, MCT estimates lifetime, performance and energy consumption using lightweight online predictors (eg. quadratic regression and gradient boosting) and a small set of congurations guided by the selected features. Finally, given the estimation of all congurations, MCT selects the optimal conguration based on the user-dened objectives. As a proof of concept, we test MCT's ability to guarantee dierent lifetime targets and achieve 95% of maximum performance, while minimizing energy consumption. We nd that MCT improves performance by 9.24% and reduces energy by 7.95% compared to the best static conguration. Moreover, the performance of MCT is 94.49% of the ideal conguration with only 5.3% more energy consumption.
This thesis is about using statistical methods for performance and power estimation which would a... more This thesis is about using statistical methods for performance and power estimation which would allow us to develop better scheduling algorithms and also more energy efficient systems. In many deployments, computer systems are underutilized – meaning that applications have performance requirements that demand less than full system capacity. Ideally, we would take advantage of this under-utilization by allocating system resources so that the performance requirements are met and energy is minimized. This optimization problem is complicated by the fact that the performance and power consumption of various system configurations are often application – or even input – dependent. Thus, practically, minimizing energy for a performance constraint requires fast, accurate estimations of application-dependent performance and power tradeoffs. We propose a set of algorithms for different scenarios to tackle this problem. First, we propose LEO, a probabilistic graphical model-based learning syste...
Participation of less-represented groups in clinical research has recently gained a great deal of... more Participation of less-represented groups in clinical research has recently gained a great deal of interest from organizations, as they have begun to recognize the importance of ensuring the safety of treatments and medications on a variety of people prior to use with the general population. One such group with the lowest amount of participation in clinical studies is minority women, as there are fewer minorities than non-minorities participating, and less women than men. One reason that has been suggested for why this occurs is that minorities are less trusting of doctors and the healthcare system. The purpose of this study in general was to determine the factors encouraging or discouraging women from clinical research, but the focus of this is on whether distrust is the root of the problem. This was accomplished by surveying women about this and other factors that may motivate their decisions to partake in clinical studies at a variety of locations, and performing statistical analy...
2017 IEEE International Conference on Autonomic Computing (ICAC), 2017
Independent applications co-scheduled on the same hardware will interfere with one another, affec... more Independent applications co-scheduled on the same hardware will interfere with one another, affecting performance in complicated ways. Predicting this interference is key to efficiently scheduling applications on shared hardware, but forming accurate predictions is difficult because there are many shared hardware features that could lead to the interference. In this paper we investigate machine learning approaches (specifically, regularization) to understand the relation between those hardware features and application interference. We propose ESP, a highly accurate and fast regularization technique for application interference prediction. To demonstrate this practicality, we implement ESP and integrate it into a scheduler for both single and multi-node Linux/x86 systems and compare the scheduling performance to state-of-the-art heuristics. We find that ESP-based schedulers increase throughput by 1.25-1.8× depending on the scheduling scenario. Additionally, we find that ESP's accurate predictions allow schedulers to avoid catastrophic decisions, which heuristic approaches fundamentally cannot detect.
New Insights into the Future of Pharmacoepidemiology and Drug Safety, 2021
The pharmacovigilance has been evolved as a professional and ethical practice in ensuring the saf... more The pharmacovigilance has been evolved as a professional and ethical practice in ensuring the safety of medicines. The Adverse Drug Reactions (ADRs) associated with the use of medicines including Anti-Tuberculous Therapy (ATT) through a robust system of pharmacovigilance helps in promoting the safety of patients at large. The occurrence of ADRs associated with the use of ATT is expected, a large number of medicines are combined and used for prolonged duration. The suspected ADRs associated with first line ATT are well documented. However, the drugs used in second line or multidrug resistant to tuberculosis (TB), namely bedaquiline, reported to cause QT prolongation in electrocardiogram reading as one of the most common ADRs. Therefore, early identification and prevention of ADRs during ATT is essential for promoting the rational use and reduce the burden of anti-microbial resistance, besides achieving better treatment outcomes.
Proceedings of the 46th International Symposium on Computer Architecture, 2019
Machine learning and artiicial intelligence are invaluable for computer systems optimization: as ... more Machine learning and artiicial intelligence are invaluable for computer systems optimization: as computer systems expose more resources for management, ML/AI is necessary for modeling these resources' complex interactions. The standard way to incorporate ML/AI into a computer system is to irst train a learner to accurately predict the system's behavior as a function of resource usageÐe.g., to predict energy eiciency as a function of core usageÐand then deploy the learned model as part of a systemÐe.g., a scheduler. In this paper, we show that (1) continued improvement of learning accuracy may not improve the systems result, but (2) incorporating knowledge of the systems problem into the learning process improves the systems results even though it may not improve overall accuracy. Speciically, we learn application performance and power as a function of resource usage with the systems goal of meeting latency constraints with minimal energy. We propose a novel generative model which improves learning accuracy given scarce data, and we propose a multi-phase sampling technique, which incorporates knowledge of the systems problem. Our results are both positive and negative. The generative model improves accuracy, even for state-of-the-art learning systems, but negatively impacts energy. Multi-phase sampling reduces energy consumption compared to the state-of-the-art, but does not improve accuracy. These results imply that learning for systems optimization may have reached a point of diminishing returns where accuracy improvements have little efect on the systems outcome. Thus we advocate that future work on learning for systems should de-emphasize accuracy and instead incorporate the system problem's structure into the learner. CCS CONCEPTS • Computing methodologies → Machine learning; • Computer systems organization → Heterogeneous (hybrid) systems; Embedded systems; Real-time system architecture; • Hardware → Chip-level power issues.
The binding free energy calculation of protein–ligand complexes is necessary for research into vi... more The binding free energy calculation of protein–ligand complexes is necessary for research into virus–host interactions and the relevant applications in drug discovery. However, many current computational methods of such calculations are either inefficient or inaccurate in practice. Utilizing implicit solvent models in the molecular mechanics generalized Born surface area (MM/GBSA) framework allows for efficient calculations without significant loss of accuracy. Here, GBNSR6, a new flavor of the generalized Born model, is employed in the MM/GBSA framework for measuring the binding affinity between SARS-CoV-2 spike protein and the human ACE2 receptor. A computational protocol is developed based on the widely studied Ras–Raf complex, which has similar binding free energy to SARS-CoV-2/ACE2. Two options for representing the dielectric boundary of the complexes are evaluated: one based on the standard Bondi radii and the other based on a newly developed set of atomic radii (OPT1), optimi...
New Insights into the Future of Pharmacoepidemiology and Drug Safety [Working Title], 2021
Medical devices are being used in healthcare facilities for diagnosis, monitoring, prevention and... more Medical devices are being used in healthcare facilities for diagnosis, monitoring, prevention and treatment of an array of diseases. To ensure user/patient safety associated with the medical devices being used in healthcare industry, it is of utmost importance to closely monitor the adverse events associated with the medical devices through a robust, sustainable and scaled surveillance. Materiovigilance Programme of India (MvPI) provides a reliable system to report adverse events associated with medical devices. Under MvPI, various modalities to report adverse events associated with medical devices have been developed. These modalities include an editable medical device adverse event reporting form, a toll-free helpline number and a field safety corrective action form (FSCA). FSCA form is used to notify the regulatory authority and healthcare professionals on corrective actions or recall by the manufacturer. Due to the emergence of the Coronavirus disease 2019 (COVID-19) pandemic, o...
Cognitive radio (CR) is a flexible wireless network that can solve the scarcity and underutilizat... more Cognitive radio (CR) is a flexible wireless network that can solve the scarcity and underutilization problem of the spectrum by permitting unlicensed users to access licensed bands. The dynamic nature of CR makes it more vulnerable in terms of security. This paper's emphasis is on the primary user emulation attack, which poses a severe threat to the spectrum sensing operation of CR. In this attack, a malicious user imitates the signal characteristics of a licensed user (primary user) to disguise its true identity. Although many survey papers enhance our knowledge of cognitive radio security, this paper is an attempt to culminate new findings with the old ones to keep up the pace of the research community. Finally, the paper summarizes with some recommendations and future strategies pertinent to energy-efficient and flexible security methods for next-generation wireless systems.
Background: The relationship between histological findings of preimplantation kidney biopsies and... more Background: The relationship between histological findings of preimplantation kidney biopsies and pump resistive index (RI) values during machine preservation and allograft survival remains unclear due to conflicting results between different studies. Objective: To determine the association between histological findings of preimplantation kidney biopsies and pump resistive index values and short-term graft survival (up to 3 years post-transplant). Methods: We performed a retrospective review of all patients who received a kidney-only transplant between January 1, 2009, and December 31, 2014, at our center and followed-up until December 1, 2016. Patients who died with a functioning allograft were excluded from analysis. Obsolete glomerulosclerosis was divided into 3 groups (< 10%, 10-20%, and > 20%), intimal arterial narrowing was divided into 3 groups (< 10%, 11-25% and > 25%) and RI was divided into 5 groups (0.00-0.09, 0.10-0.19, 0.20-0.29, 0.30-0.39 and 0.40=0.49). The log-rank test was used to test if there is a significant difference in different RI groups in terms of the probability of survival.
Uploads
Papers by Nikita mishra