Land degradation is one of the major threats faced globally, affecting arid and other dryland are... more Land degradation is one of the major threats faced globally, affecting arid and other dryland areas severely. In such areas, biomass productivity is a function of precipitation to a large extent. This paper discusses methodology developed to generate land productivity dynamics assessment map using vegetation precipitation relationship residuals as proxy parameter to land degradation using long-term time series of satellite-derived Normalised Vegetation Difference Index and climate model-based precipitation data. This developed methodology rationally utilises information retrieved from statistical methods as well as qualitative measures. Application of developed methodology indicates that 62.54% of total land area of Rajasthan state is showing symptoms of increasing land productivity as compared to 2.35% area of decreasing land productivity.
Proceedings of the AAAI Conference on Artificial Intelligence
Much of the focus on finding good representations in reinforcement learning has been on learning ... more Much of the focus on finding good representations in reinforcement learning has been on learning complex non-linear predictors of value. Policy gradient algorithms, which directly represent the policy, often need fewer parameters to learn good policies. However, they typically employ a fixed parametric representation that may not be sufficient for complex domains. This paper introduces the Policy Tree algorithm, which can learn an adaptive representation of policy in the form of a decision tree over different instantiations of a base policy. Policy gradient is used both to optimize the parameters and to grow the tree by choosing splits that enable the maximum local increase in the expected return of the policy. Experiments show that this algorithm can choose genuinely helpful splits and significantly improve upon the commonly used linear Gibbs softmax policy, which we choose as our base policy.
Accurate diagnosis is crucial for preventing the progression of Parkinson's, as well as impro... more Accurate diagnosis is crucial for preventing the progression of Parkinson's, as well as improving the quality of life with individuals with Parkinson's disease. In this paper, we develop a gender specific and age dependent classification method to diagnose the Parkinson's disease using the handwriting based measurements. The gender specific and age dependent classifier was observed significantly outperforming the generalized classifier. An improved accuracy of 83.75% (SD=1.63) with the female specific classifier, and 79.55% (SD=1.58) with the old age dependent classifier was observed in comparison to 75.76% (SD=1.17) accuracy with the generalized classifier. Finally, combining the age and gender information proved to be encouraging in classification. We performed a rigorous analysis to observe the dominance of gender specific and age dependent features for Parkinson's detection and ranked them using the support vector machine(SVM) ranking method. Distinct set of feat...
Proceedings of the International Conference on Computer-Aided Design, 2018
Energy efficiency and performance of heterogeneous multiprocessor systems-on-chip (SoC) depend cr... more Energy efficiency and performance of heterogeneous multiprocessor systems-on-chip (SoC) depend critically on utilizing a diverse set of processing elements and managing their power states dynamically. Dynamic resource management techniques typically rely on power consumption and performance models to assess the impact of dynamic decisions. Despite the importance of these decisions, many existing approaches rely on fixed power and performance models learned offline. This paper presents an online learning framework to construct adaptive analytical models. We illustrate this framework for modeling GPU frame processing time, GPU power consumption and SoC power-temperature dynamics. Experiments on Intel Atom E3826, Qualcomm Snapdragon 810, and Samsung Exynos 5422 SoCs demonstrate that the proposed approach achieves less than 6% error under dynamically varying workloads.
Accurate diagnosis is crucial for preventing the progression of Parkinson's, as well as impro... more Accurate diagnosis is crucial for preventing the progression of Parkinson's, as well as improving the quality of life with individuals with Parkinson's disease. In this paper, we develop a sex-specific and age-dependent classification method to diagnose the Parkinson's disease using the online handwriting recorded from individuals with Parkinson's(n=37;m/f-19/18;age-69.3+-10.9years) and healthy controls(n=38;m/f-20/18;age-62.4+-11.3 years).The sex specific and age dependent classifier was observed significantly outperforming the generalized classifier. An improved accuracy of 83.75%(SD+1.63) with female specific classifier, and 79.55%(SD=1.58) with old age dependent classifier was observed in comparison to 75.76%(SD=1.17) accuracy with the generalized classifier. Finally, combining the age and sex information proved to be encouraging in classification. We performed a rigorous analysis to observe the dominance of sex specific and age dependent features for Parkinson&#...
The enhancement is required in every field. Day by day our world is getting modernized. We need t... more The enhancement is required in every field. Day by day our world is getting modernized. We need to adapt new technology to survive in the world of competitors. So people try to introduce the new technology and outdate the previous one. Similarly in the case of networks we make enhanced versions of networks. 4G is faster than 3G network with less buffering, better audio quality, streaming services with reduced lag and improved gaming experience. The evolution of 5G networks will make our lives even more easier. 5G technology will work on sustainability , efficiency. It will be using space-saving approach. It will also help in minimising the energy requirements of web device and network infrastructure. Keywords5G,CDMA,TDM,SMS.
International Journal for Research in Applied Science and Engineering Technology, 2020
As technology changes, it becomes increasingly challenging for businesses of all types to keep th... more As technology changes, it becomes increasingly challenging for businesses of all types to keep their personal and customer's information on the web secure. Web security is important to keeping hackers and cyber-thieves from accessing sensitive information. Without a proactive security strategy, businesses risk the spread and escalation of malware, attacks on other websites, networks, and other IT infrastructures. If a hacker is successful, attacks can spread from computer to computer, making it difficult to find the origin. This project deals with preventing the potential errors while developing a basic website in order to prevent it from possible cyber-attacks. Cyber-attacks will be performed on unsecured site and then its vulnerabilities will be compared with the secured site.
Much of the focus on finding good representations in reinforcement learning has been on learning ... more Much of the focus on finding good representations in reinforcement learning has been on learning complex non-linear predictors of value. Methods like policy gradient, that do not learn a value function and instead directly represent policy, often need fewer parameters to learn good policies. However, they typically employ a fixed parametric representation that may not be sufficient for complex domains.
The aim is to achieve low-cost, wireless voice data transmission and develop an intelligent netwo... more The aim is to achieve low-cost, wireless voice data transmission and develop an intelligent network of transmitters for tracking vehicles enabled by Traffic Auditory signal hearing aid (T-Asha) devices which are fitted inside special crash helmets. A Texas Instruments (TI) microcontroller (MCU), MSP430 series, is used to reproduce encoded audio information.
Wearable devices have the potential to transform multiple facets of human life, including healthc... more Wearable devices have the potential to transform multiple facets of human life, including healthcare, activity monitoring, and interaction with computers. However, a number of technical and adaptation challenges hinder the widespread and daily usage of wearable devices. Recent research efforts have focused on identifying these challenges and solving them such that the potential of wearable devices can be realized. This monograph starts with a survey of the recent literature on the challenges faced by wearable devices. Then, it discusses potential solutions to each of the challenges. We start with the primary application areas that provide value to the users of wearable devices. We then present recent work on the design of physically flexible and bendable
Thrombotic Thrombocytopenic Purpura (TTP), first described in the early twentieth century, was fa... more Thrombotic Thrombocytopenic Purpura (TTP), first described in the early twentieth century, was fatal in the majority of cases until the advent of plasma therapy. Mortality has declined dramatically since the 1980s, and in the 1990s the pathophysiology was elucidated through the discovery and understanding of the central role of a deficiency of the metalloprotease ADAMTS13. The vast majority of cases occur due to an autoimmune process, with a minority of cases due to an underlying mutation. Since the turn of this century, Rituximab, a monoclonal antibody targeting CD20 expressed on B-lymphocytes, has become widely used to treat patients with a wide spectrum of autoimmune diseases, including TTP. However, Rituximab remains "off-label" in the setting of TTP. We recently encountered a patient with chronic relapsing TTP, whose clinical relapses have responded to Rituximabbased therapy without need to resort to plasmapheresis. The clinical course of this patient is described, and the literature focusing on the use of Rituximab in prophylaxis to prevent recurrent TTP is reviewed.
Computer Methods and Programs in Biomedicine, 2019
Background and Objectives: Diagnosis of Parkinson's with higher accuracy is always desirable to s... more Background and Objectives: Diagnosis of Parkinson's with higher accuracy is always desirable to slow down the progression of the disease and improved quality of life. There are evidences of inherent neurological differences between male and females as well as between elderly and adults. However, the potential of such gender and age infomration have not been exploited yet for Parkinson's identification. Methods: In this paper, we develop a sex-specific and age-dependent classification method to diagnose the Parkinson's disease using the online handwriting recorded from individuals with Parkinson's (n = 37; m/f-19/18;age-69.3±10.9yrs) and healthy controls (n = 38; m/f-20/18;age-62.4±11.3yrs). A support vector machine ranking method is used to present the features specific to their dominance in sex and age group for Parkinson's diagnosis. Results: The sex-specific and age-dependent classifier was observed significantly outperforming the generalized classifier. An improved accuracy of 83.75% (SD = 1.63) with the female-specific classifier, and 79.55% (SD = 1.58) with the old-age dependent classifier was observed in comparison to 75.76% (SD = 1.17) accuracy with the generalized classifier. Conclusions: Combining the age and sex information proved to be encouraging in classification. A distinct set of features were observed to be dominating for higher classification accuracy in a different category of classification.
IEEE Transactions on Multi-Scale Computing Systems, 2017
Mechanically flexible, printed and stretchable electronics are gaining momentum. Rapid progress a... more Mechanically flexible, printed and stretchable electronics are gaining momentum. Rapid progress at device and circuit levels are already underway, but researchers are yet to envision the system design in a flexible form. This paper introduces Systemson-Polymer (SoP) based on flexible hybrid electronics (FHE) to combine the advantages of flexible electronics and traditional silicon technology. First, we formally define flexibility as a new design metric in addition to existing power, performance, and area metrics. Then, we present a novel optimization approach to place rigid components onto a flexible substrate while minimizing the loss in flexibility. We show that the optimal placement leads to as much as 5.7× enhancement in flexibility compared to the naïve placement. We confirm the accuracy of our models and optimization framework using a finite element method (FEM) simulator. Finally, we demonstrate the SoP concept using a concrete hardware prototype and discuss the major challenges in the architecture and design of SoPs.
Heterogeneous multiprocessor system-on-chips (SoCs) provide a wide range of parameters that can b... more Heterogeneous multiprocessor system-on-chips (SoCs) provide a wide range of parameters that can be managed dynamically. For example, one can control the type (big/little), number and frequency of active cores in state-of-the-art mobile processors at runtime. These runtime choices lead to more than 10× range in execution time, 5× range in power consumption, and 50× range in performance per watt. Therefore, it is crucial to make optimum power management decisions as a function of dynamically varying workloads at runtime. This paper presents a reinforcement learning approach for dynamically controlling the number and frequency of active big and little cores in mobile processors. We propose an efficient deep Q-learning methodology to optimize the performance per watt (PPW). Experiments using Odroid XU3 mobile platform show that the PPW achieved by the proposed approach is within 1% of the optimal value obtained by an oracle.
Approximately 18% of the 3.2 million smartphone applications rely on integrated graphics processi... more Approximately 18% of the 3.2 million smartphone applications rely on integrated graphics processing units (GPUs) to achieve competitive performance. Graphics performance, typically measured in frames per second, is a strong function of the GPU frequency, which in turn has a significant impact on mobile processor power consumption. Consequently, dynamic power management algorithms have to assess the performance sensitivity to the frequency accurately to choose the operating frequency of the GPU effectively. Since the impact of GPU frequency on performance varies rapidly over time, there is a need for online performance models that can adapt to varying workloads. This paper presents a lightweight adaptive runtime performance model that predicts the frame processing time of graphics workloads at runtime without apriori characterization. We employ this model to estimate the frame time sensitivity to the GPU frequency, i.e., the partial derivative of the frame time with respect to the GPU frequency. The proposed model does not rely on any parameter learned offline. Our experiments on the Intel Minnowboard MAX platform running common GPU benchmarks show that the mean absolute percentage error in frame time and frame time sensitivity prediction are 4.2% and 6.7%, respectively.
Land degradation is one of the major threats faced globally, affecting arid and other dryland are... more Land degradation is one of the major threats faced globally, affecting arid and other dryland areas severely. In such areas, biomass productivity is a function of precipitation to a large extent. This paper discusses methodology developed to generate land productivity dynamics assessment map using vegetation precipitation relationship residuals as proxy parameter to land degradation using long-term time series of satellite-derived Normalised Vegetation Difference Index and climate model-based precipitation data. This developed methodology rationally utilises information retrieved from statistical methods as well as qualitative measures. Application of developed methodology indicates that 62.54% of total land area of Rajasthan state is showing symptoms of increasing land productivity as compared to 2.35% area of decreasing land productivity.
Proceedings of the AAAI Conference on Artificial Intelligence
Much of the focus on finding good representations in reinforcement learning has been on learning ... more Much of the focus on finding good representations in reinforcement learning has been on learning complex non-linear predictors of value. Policy gradient algorithms, which directly represent the policy, often need fewer parameters to learn good policies. However, they typically employ a fixed parametric representation that may not be sufficient for complex domains. This paper introduces the Policy Tree algorithm, which can learn an adaptive representation of policy in the form of a decision tree over different instantiations of a base policy. Policy gradient is used both to optimize the parameters and to grow the tree by choosing splits that enable the maximum local increase in the expected return of the policy. Experiments show that this algorithm can choose genuinely helpful splits and significantly improve upon the commonly used linear Gibbs softmax policy, which we choose as our base policy.
Accurate diagnosis is crucial for preventing the progression of Parkinson's, as well as impro... more Accurate diagnosis is crucial for preventing the progression of Parkinson's, as well as improving the quality of life with individuals with Parkinson's disease. In this paper, we develop a gender specific and age dependent classification method to diagnose the Parkinson's disease using the handwriting based measurements. The gender specific and age dependent classifier was observed significantly outperforming the generalized classifier. An improved accuracy of 83.75% (SD=1.63) with the female specific classifier, and 79.55% (SD=1.58) with the old age dependent classifier was observed in comparison to 75.76% (SD=1.17) accuracy with the generalized classifier. Finally, combining the age and gender information proved to be encouraging in classification. We performed a rigorous analysis to observe the dominance of gender specific and age dependent features for Parkinson's detection and ranked them using the support vector machine(SVM) ranking method. Distinct set of feat...
Proceedings of the International Conference on Computer-Aided Design, 2018
Energy efficiency and performance of heterogeneous multiprocessor systems-on-chip (SoC) depend cr... more Energy efficiency and performance of heterogeneous multiprocessor systems-on-chip (SoC) depend critically on utilizing a diverse set of processing elements and managing their power states dynamically. Dynamic resource management techniques typically rely on power consumption and performance models to assess the impact of dynamic decisions. Despite the importance of these decisions, many existing approaches rely on fixed power and performance models learned offline. This paper presents an online learning framework to construct adaptive analytical models. We illustrate this framework for modeling GPU frame processing time, GPU power consumption and SoC power-temperature dynamics. Experiments on Intel Atom E3826, Qualcomm Snapdragon 810, and Samsung Exynos 5422 SoCs demonstrate that the proposed approach achieves less than 6% error under dynamically varying workloads.
Accurate diagnosis is crucial for preventing the progression of Parkinson's, as well as impro... more Accurate diagnosis is crucial for preventing the progression of Parkinson's, as well as improving the quality of life with individuals with Parkinson's disease. In this paper, we develop a sex-specific and age-dependent classification method to diagnose the Parkinson's disease using the online handwriting recorded from individuals with Parkinson's(n=37;m/f-19/18;age-69.3+-10.9years) and healthy controls(n=38;m/f-20/18;age-62.4+-11.3 years).The sex specific and age dependent classifier was observed significantly outperforming the generalized classifier. An improved accuracy of 83.75%(SD+1.63) with female specific classifier, and 79.55%(SD=1.58) with old age dependent classifier was observed in comparison to 75.76%(SD=1.17) accuracy with the generalized classifier. Finally, combining the age and sex information proved to be encouraging in classification. We performed a rigorous analysis to observe the dominance of sex specific and age dependent features for Parkinson&#...
The enhancement is required in every field. Day by day our world is getting modernized. We need t... more The enhancement is required in every field. Day by day our world is getting modernized. We need to adapt new technology to survive in the world of competitors. So people try to introduce the new technology and outdate the previous one. Similarly in the case of networks we make enhanced versions of networks. 4G is faster than 3G network with less buffering, better audio quality, streaming services with reduced lag and improved gaming experience. The evolution of 5G networks will make our lives even more easier. 5G technology will work on sustainability , efficiency. It will be using space-saving approach. It will also help in minimising the energy requirements of web device and network infrastructure. Keywords5G,CDMA,TDM,SMS.
International Journal for Research in Applied Science and Engineering Technology, 2020
As technology changes, it becomes increasingly challenging for businesses of all types to keep th... more As technology changes, it becomes increasingly challenging for businesses of all types to keep their personal and customer's information on the web secure. Web security is important to keeping hackers and cyber-thieves from accessing sensitive information. Without a proactive security strategy, businesses risk the spread and escalation of malware, attacks on other websites, networks, and other IT infrastructures. If a hacker is successful, attacks can spread from computer to computer, making it difficult to find the origin. This project deals with preventing the potential errors while developing a basic website in order to prevent it from possible cyber-attacks. Cyber-attacks will be performed on unsecured site and then its vulnerabilities will be compared with the secured site.
Much of the focus on finding good representations in reinforcement learning has been on learning ... more Much of the focus on finding good representations in reinforcement learning has been on learning complex non-linear predictors of value. Methods like policy gradient, that do not learn a value function and instead directly represent policy, often need fewer parameters to learn good policies. However, they typically employ a fixed parametric representation that may not be sufficient for complex domains.
The aim is to achieve low-cost, wireless voice data transmission and develop an intelligent netwo... more The aim is to achieve low-cost, wireless voice data transmission and develop an intelligent network of transmitters for tracking vehicles enabled by Traffic Auditory signal hearing aid (T-Asha) devices which are fitted inside special crash helmets. A Texas Instruments (TI) microcontroller (MCU), MSP430 series, is used to reproduce encoded audio information.
Wearable devices have the potential to transform multiple facets of human life, including healthc... more Wearable devices have the potential to transform multiple facets of human life, including healthcare, activity monitoring, and interaction with computers. However, a number of technical and adaptation challenges hinder the widespread and daily usage of wearable devices. Recent research efforts have focused on identifying these challenges and solving them such that the potential of wearable devices can be realized. This monograph starts with a survey of the recent literature on the challenges faced by wearable devices. Then, it discusses potential solutions to each of the challenges. We start with the primary application areas that provide value to the users of wearable devices. We then present recent work on the design of physically flexible and bendable
Thrombotic Thrombocytopenic Purpura (TTP), first described in the early twentieth century, was fa... more Thrombotic Thrombocytopenic Purpura (TTP), first described in the early twentieth century, was fatal in the majority of cases until the advent of plasma therapy. Mortality has declined dramatically since the 1980s, and in the 1990s the pathophysiology was elucidated through the discovery and understanding of the central role of a deficiency of the metalloprotease ADAMTS13. The vast majority of cases occur due to an autoimmune process, with a minority of cases due to an underlying mutation. Since the turn of this century, Rituximab, a monoclonal antibody targeting CD20 expressed on B-lymphocytes, has become widely used to treat patients with a wide spectrum of autoimmune diseases, including TTP. However, Rituximab remains "off-label" in the setting of TTP. We recently encountered a patient with chronic relapsing TTP, whose clinical relapses have responded to Rituximabbased therapy without need to resort to plasmapheresis. The clinical course of this patient is described, and the literature focusing on the use of Rituximab in prophylaxis to prevent recurrent TTP is reviewed.
Computer Methods and Programs in Biomedicine, 2019
Background and Objectives: Diagnosis of Parkinson's with higher accuracy is always desirable to s... more Background and Objectives: Diagnosis of Parkinson's with higher accuracy is always desirable to slow down the progression of the disease and improved quality of life. There are evidences of inherent neurological differences between male and females as well as between elderly and adults. However, the potential of such gender and age infomration have not been exploited yet for Parkinson's identification. Methods: In this paper, we develop a sex-specific and age-dependent classification method to diagnose the Parkinson's disease using the online handwriting recorded from individuals with Parkinson's (n = 37; m/f-19/18;age-69.3±10.9yrs) and healthy controls (n = 38; m/f-20/18;age-62.4±11.3yrs). A support vector machine ranking method is used to present the features specific to their dominance in sex and age group for Parkinson's diagnosis. Results: The sex-specific and age-dependent classifier was observed significantly outperforming the generalized classifier. An improved accuracy of 83.75% (SD = 1.63) with the female-specific classifier, and 79.55% (SD = 1.58) with the old-age dependent classifier was observed in comparison to 75.76% (SD = 1.17) accuracy with the generalized classifier. Conclusions: Combining the age and sex information proved to be encouraging in classification. A distinct set of features were observed to be dominating for higher classification accuracy in a different category of classification.
IEEE Transactions on Multi-Scale Computing Systems, 2017
Mechanically flexible, printed and stretchable electronics are gaining momentum. Rapid progress a... more Mechanically flexible, printed and stretchable electronics are gaining momentum. Rapid progress at device and circuit levels are already underway, but researchers are yet to envision the system design in a flexible form. This paper introduces Systemson-Polymer (SoP) based on flexible hybrid electronics (FHE) to combine the advantages of flexible electronics and traditional silicon technology. First, we formally define flexibility as a new design metric in addition to existing power, performance, and area metrics. Then, we present a novel optimization approach to place rigid components onto a flexible substrate while minimizing the loss in flexibility. We show that the optimal placement leads to as much as 5.7× enhancement in flexibility compared to the naïve placement. We confirm the accuracy of our models and optimization framework using a finite element method (FEM) simulator. Finally, we demonstrate the SoP concept using a concrete hardware prototype and discuss the major challenges in the architecture and design of SoPs.
Heterogeneous multiprocessor system-on-chips (SoCs) provide a wide range of parameters that can b... more Heterogeneous multiprocessor system-on-chips (SoCs) provide a wide range of parameters that can be managed dynamically. For example, one can control the type (big/little), number and frequency of active cores in state-of-the-art mobile processors at runtime. These runtime choices lead to more than 10× range in execution time, 5× range in power consumption, and 50× range in performance per watt. Therefore, it is crucial to make optimum power management decisions as a function of dynamically varying workloads at runtime. This paper presents a reinforcement learning approach for dynamically controlling the number and frequency of active big and little cores in mobile processors. We propose an efficient deep Q-learning methodology to optimize the performance per watt (PPW). Experiments using Odroid XU3 mobile platform show that the PPW achieved by the proposed approach is within 1% of the optimal value obtained by an oracle.
Approximately 18% of the 3.2 million smartphone applications rely on integrated graphics processi... more Approximately 18% of the 3.2 million smartphone applications rely on integrated graphics processing units (GPUs) to achieve competitive performance. Graphics performance, typically measured in frames per second, is a strong function of the GPU frequency, which in turn has a significant impact on mobile processor power consumption. Consequently, dynamic power management algorithms have to assess the performance sensitivity to the frequency accurately to choose the operating frequency of the GPU effectively. Since the impact of GPU frequency on performance varies rapidly over time, there is a need for online performance models that can adapt to varying workloads. This paper presents a lightweight adaptive runtime performance model that predicts the frame processing time of graphics workloads at runtime without apriori characterization. We employ this model to estimate the frame time sensitivity to the GPU frequency, i.e., the partial derivative of the frame time with respect to the GPU frequency. The proposed model does not rely on any parameter learned offline. Our experiments on the Intel Minnowboard MAX platform running common GPU benchmarks show that the mean absolute percentage error in frame time and frame time sensitivity prediction are 4.2% and 6.7%, respectively.
Uploads
Papers by Ujjwal gupta