In this paper, the system including communicable infection from prey to predator, represented by ... more In this paper, the system including communicable infection from prey to predator, represented by the growing rate of the prey as a healing of a predator's, has been utilized to develop and analyze a four-dimensional, non-linear eco-epidemic model. In the context of the analysis, the exploration of all potential equilibrium points, along with an examination of their local stability conditions, has been conducted both with and without considering time delays. Additionally, the confirmation of the model's positivity and boundedness has been carried out. The positive and negative impact of the proposed model on the prey-predator population has been investigated aided by sensitivity analysis where the presence and validence of Bifurcation were elucidated by the numerical studies. The parameters were identified which explained the influence of disease and recovery delay over the model population. A right base for the perception of the behavioural effects of the prey-predator on eco-epidemiology has been prepared through the theoretical result. Based on the analytical result furthermore, numerical simulations were done so that the authenticity of the author's numerical analytical approach using parameter values may be verified.
Extremists are increasingly using social media to recruit and radicalize other users and increase... more Extremists are increasingly using social media to recruit and radicalize other users and increase their money. Terrorists can use popular social networks accounts and perform their activities in a hidden way. So, it is crucial to create a fruitful mechanism for controlling the spread of misinformation. Otherwise, a large number of people can mislead by this terrorist activity by joining them. Here, we propose malicious news spreading model incorporating hidden attackers of a social network. A threshold is defined for deciding the extinction of malicious news from a social network. Here, we show the importance of network alertness and activity of cyber-security agencies in the modified model. Moreover, we obtained the optimal values of the control parameters for emergencies.
International Journal of Biomedical Engineering and Technology, 2015
The main goal of the research work is to apply a Genetic Algorithm (GA) in order to prune the inp... more The main goal of the research work is to apply a Genetic Algorithm (GA) in order to prune the inputs for an Artificial Neural Network (ANN) for medical diagnosis in order to reduce the computational complexity. The inputs in medical diagnosis are the diagnostic factors. The GA implemented creates the essential and minimal subset of diagnostic factors required for medical diagnosis. Firstly, the ANN is applied alone and the time taken and efficiency of the medical diagnostic system are recorded. Then, pruning of inputs using GA and then the pruned inputs are used for the ANN, and the time taken and efficiency obtained are compared with the previous one. The medical diagnostic data set is taken from UCI medical repository for the hepatitis disease. There is a significant percentage reduction in training time as well as testing time of ANN and a significant improvement in the success rate of diagnosis.
We describe a new approach for face recognition using modular neural networks with a fuzzy logic ... more We describe a new approach for face recognition using modular neural networks with a fuzzy logic method for response integration. We proposed a new architecture for modular neural net-works for achieving pattern recognition in the particular case of human faces. Also, ...
This article investigates a tri-trophic food chain model with the epidemic at the bottom level (p... more This article investigates a tri-trophic food chain model with the epidemic at the bottom level (prey) and gestation delays for the next two levels( the intermediate predator and top predator). The steady states' positivity, boundedness, existence, and stability have been analyzed. The study of Hopf bifurcation has been carried out for different steady states. It is shown that when the gestation delay for intermediate predator τ 1 is absent, the interior equilibrium is locally stable for gestation delays of top predators τ 2 ∈ (0, τ + 20 ) and performs Hopf bifurcation for all τ 2 > τ + 20 . Further, in the presence both the delay τ 1 and τ 2 , the system becomes unstable when τ 1 , τ 2 crosses their critical values τ ++ 10 , τ ++ 20 respectively and executes Hopf bifurcation τ 1 > τ ++ 10 , τ 2 > τ ++ 20 . Moreover, the proposed system is stable for τ 1 > τ ++ 10 , τ 2 < τ ++ 20 and unstable for τ 1 < τ ++ 10 , τ 2 > τ ++ 20 . The sensitivity analysis of the system at coexistence steady state are presented to identify the effect of system parameters on the variables. Finally, we have presented a numerical simulation to verify our analytical findings.
Deep Learning has made great impact on computer vision applications such as object detection, ima... more Deep Learning has made great impact on computer vision applications such as object detection, image classification, segmentation and localization. Computer based system for lung cancer detection and diagnosis helps in detecting the lung nodules and analysing the nodule features for further segmentation and classification. This paper explains the effect of UNet model on various back bone architectures pretrained on ImageNet dataset. This work was carried out on LIDC-IDRI dataset and the efficiency interms of IoU score and accuracy both in training and validation datasets are listed. Results show that EfficientNet backbone outperforms other backbone architectures.
kNN is a very effective Instance based learning method, and it is easy to implement. Due to heter... more kNN is a very effective Instance based learning method, and it is easy to implement. Due to heterogeneous nature of data, noises from different possible sources are also widespread in nature especially in case of large-scale databases. For noise elimination and effect of pseudo neighbours, in this paper, we propose a new learning algorithm which performs the task of anomaly detection and removal of pseudo neighbours from the dataset so as to provide comparative better results. This algorithm also tries to minimize effect of those neighbours which are distant. A concept of certainty measure is also introduced for experimental results. The advantage of using concept of mutual neighbours and distance-weighted voting is that, dataset will be refined after removal of anomaly and weightage concept compels to take into account more consideration of those neighbours, which are closer. Consequently, finally the performance of proposed algorithm is calculated.
Communications in computer and information science, 2020
Image based Lung cancer diagnosis offers an effective and non-intrusive method for identification... more Image based Lung cancer diagnosis offers an effective and non-intrusive method for identification and diagnosis of lung cancer in a patient. This is done by segmentation of lung nodules from radiological images obtained from patients. This paper proposes a framework to segment nodule lesions in images. The novelty of the proposed framework is that we use optical flow information along with the morphological operations on images. The proposed framework applies the Farneback method of optical flow extraction algorithms for the nodule segmentation task. This work validates and demonstrates that optical flow information is useful for nodule segmentation and can be used in tandom with other features.
The Plant Pest problem is a crucial area of investigation for the present time. Myriad chemical s... more The Plant Pest problem is a crucial area of investigation for the present time. Myriad chemical solutions are available to solve this issue, but due to their toxic outcomes, the Natural enemy of pests is come across to be a better solution. This article proposes a fractional calculus approach to a plant pest mature and immature Natural enemy model. Existence and uniqueness of the solution and non-negativity of the solution is being discussed also reached on the Global and Local stability of the equilibrium point. As memory is always present in biological system that is the reason the suggested technique is more advantageous due to the memory effect of fractional derivatives and produces more appropriate solutions. Simulation study for different fractional parameters generates the optimum solution.
In this paper, a non-linear mathematical model with three irresistible classes for the impacts of... more In this paper, a non-linear mathematical model with three irresistible classes for the impacts of media awareness programs on the spread of irresistible infections, for example, influenza, has been proposed and analyzed. In the modeling process, it is expected that illness spreads because of the contact between the susceptibles and infectives, as it were. The growth rate of media awareness programs influencing the populace corresponds to the number of infective people. We examine the dynamical behavior and systematic investigation of the framework for the model, which demonstrates that the model has two equilibrium points, i.e., disease-free equilibrium (DFE) and interior (endemic) equilibrium. The outcomes show that the primary reproduction number determines the dynamics of the model. For the basic reproduction number R 0 < 1, the disease-free equilibrium is locally as well as globally asymptotically stable under a specific parameter set. If R 0 > 1, the model at the interior equilibrium is locally asymptotically stable. At long last, numerical arrangements of the model validate the analytical outcomes and facilitate a sensitivity analysis of the model parameters.
Extremists are increasingly using social media to recruit and radicalize other users and increase... more Extremists are increasingly using social media to recruit and radicalize other users and increase their money. Terrorists can use popular social networks accounts and perform their activities in a hidden way. So, it is crucial to create a fruitful mechanism for controlling the spread of misinformation. Otherwise, a large number of people can mislead by this terrorist activity by joining them. Here, we propose malicious news spreading model incorporating hidden attackers of a social network. A threshold is defined for deciding the extinction of malicious news from a social network. Here, we show the importance of network alertness and activity of cyber-security agencies in the modified model. Moreover, we obtained the optimal values of the control parameters for emergencies.
Zenodo (CERN European Organization for Nuclear Research), Apr 29, 2022
With the emergence of technologies and the increasing number of users of the Internet, such as th... more With the emergence of technologies and the increasing number of users of the Internet, such as the Internet of Things (IoT) paradigm, there are new and modern efforts to invade networks and computer systems. Many researchers have developed various artificial intelligence-based algorithms to detect these attacks. For network security they focus on machine learning models that are used in IoT and intrusion detection. In this study, we have proposed a machine learning based intrusion detection system which is a combination of support vector machine (SVM) and Genetic Algorithm (GA). GA is basically used for feature selection and parameter optimization of SVM models. The performance of a GA-based SVM model is evaluated on an KDD Cup 99 intrusion database. First, GA optimizes the KDD Cup 99 dataset and then optimizes the weight and parameters of the SVM model. The experimental result presents that the SVM model gives prominence in terms of detection rate, accuracy, false positive rate and false negative rate and also compared to other literature works.
Seasonal variability strongly affects the animal population in wildlife. It becomes essential to ... more Seasonal variability strongly affects the animal population in wildlife. It becomes essential to model seasonality in eco-epidemic dynamics to know the effect of system parameters in a periodic environment. This article presents a set of non-autonomous differential equations with time-varying disease transmission rates among prey and predators, the mortality rate of a diseased predator, the predation rate of healthy prey, and an additional food supply. The positiveness, boundedness, and presence of solution are derived. We have proved that the infection-free state is stable if periodic basic reproduction number $$R_C(t)<1$$ R C ( t ) < 1 . The stability of the coexistence state is shown at $$R_C(t)>1$$ R C ( t ) > 1 using the Poincare map and comparison theory. The significance of the parameters related to disease transmission and prevalence is described using sensitivity analysis. Numerical simulation verified our analytical findings and proved that the predator control...
The pandemic being a health issue at its core is a multifaceted crisis encompassing both economic... more The pandemic being a health issue at its core is a multifaceted crisis encompassing both economic and epidemic factors in a twisted tale of challenges. In counteraction, we have proposed a combined epidemic-economic model that analyses system dynamics arising in the presence of an infectious disease (SARS-2-COVID-19 in our case). Dynamical analysis of the system has been performed in context to the equilibria along with local and global stability analysis of the system simultaneously visualizing the effect on capital stabilization. The global stability analysis has been performed using graph-theoretic method. Curve-Fit has been performed for the system using optimization algorithm. The relation between all the parameters and variables involved in the model has been explored by calculating sensitivity indices which gives us the proportion that a relative change in a parameter brings to the relative change in a variable. Our findings reveal that (1) Vaccination instigates economic growth (with evidence of data obtained for 24 countries). ( ) Complete vaccination leads to a considerable reduction in all infections (reduction up to 90%, as per current CDC study). ( ) Excessive exposure to media can facilitate spike in infections. (4) Parameter sensitivity analysis can be of immense help in policy formation.
Proceedings of the 2018 International Conference on Algorithms, Computing and Artificial Intelligence, 2018
In this paper, we present a deep learning based approach to assign POS tags to words in a piece o... more In this paper, we present a deep learning based approach to assign POS tags to words in a piece of text given to it as input. We propose an unsupervised approach owing to the lack of a large Sanskrit annotated corpora and use the untagged Sanskrit Corpus prepared by JNU for our purpose. The only tagged corpora for Sanskrit is created by JNU which has 115,000 words which are not sufficient to apply supervised deep learning approaches. For the tag assignment purpose and determining model accuracy, we utilize this tagged corpus. We explore various methods through which each Sanskrit word can be represented as a point multi-dimensional vector space whose position accurately captures its meaning and semantic information associated with it. We also explore other data sources to improve performance and robustness of the vector representations. We use these rich vector representations and explore autoencoder based approaches for dimensionality reduction to compress these into encodings which are suitable for clustering in the vector space. We experiment with different dimensions of these compressed representations and present one which was found to offer the best clustering performance. For modelling the sequence in order to preserve the semantic information we feed these embeddings to a bidirectional LSTM autoencoder. We assign a POS tag to each of the obtained clusters and produce our result by testing the model on the tagged corpus.
kNN is a very effective Instance based learning method, and it is easy to implement. Due to heter... more kNN is a very effective Instance based learning method, and it is easy to implement. Due to heterogeneous nature of data, noises from different possible sources are also widespread in nature especially in case of large-scale databases. For noise elimination and effect of pseudo neighbours, in this paper, we propose a new learning algorithm which performs the task of anomaly detection and removal of pseudo neighbours from the dataset so as to provide comparative better results. This algorithm also tries to minimize effect of those neighbours which are distant. A concept of certainty measure is also introduced for experimental results. The advantage of using concept of mutual neighbours and distance-weighted voting is that, dataset will be refined after removal of anomaly and weightage concept compels to take into account more consideration of those neighbours, which are closer. Consequently, finally the performance of proposed algorithm is calculated.
In this paper, the system including communicable infection from prey to predator, represented by ... more In this paper, the system including communicable infection from prey to predator, represented by the growing rate of the prey as a healing of a predator's, has been utilized to develop and analyze a four-dimensional, non-linear eco-epidemic model. In the context of the analysis, the exploration of all potential equilibrium points, along with an examination of their local stability conditions, has been conducted both with and without considering time delays. Additionally, the confirmation of the model's positivity and boundedness has been carried out. The positive and negative impact of the proposed model on the prey-predator population has been investigated aided by sensitivity analysis where the presence and validence of Bifurcation were elucidated by the numerical studies. The parameters were identified which explained the influence of disease and recovery delay over the model population. A right base for the perception of the behavioural effects of the prey-predator on eco-epidemiology has been prepared through the theoretical result. Based on the analytical result furthermore, numerical simulations were done so that the authenticity of the author's numerical analytical approach using parameter values may be verified.
Extremists are increasingly using social media to recruit and radicalize other users and increase... more Extremists are increasingly using social media to recruit and radicalize other users and increase their money. Terrorists can use popular social networks accounts and perform their activities in a hidden way. So, it is crucial to create a fruitful mechanism for controlling the spread of misinformation. Otherwise, a large number of people can mislead by this terrorist activity by joining them. Here, we propose malicious news spreading model incorporating hidden attackers of a social network. A threshold is defined for deciding the extinction of malicious news from a social network. Here, we show the importance of network alertness and activity of cyber-security agencies in the modified model. Moreover, we obtained the optimal values of the control parameters for emergencies.
International Journal of Biomedical Engineering and Technology, 2015
The main goal of the research work is to apply a Genetic Algorithm (GA) in order to prune the inp... more The main goal of the research work is to apply a Genetic Algorithm (GA) in order to prune the inputs for an Artificial Neural Network (ANN) for medical diagnosis in order to reduce the computational complexity. The inputs in medical diagnosis are the diagnostic factors. The GA implemented creates the essential and minimal subset of diagnostic factors required for medical diagnosis. Firstly, the ANN is applied alone and the time taken and efficiency of the medical diagnostic system are recorded. Then, pruning of inputs using GA and then the pruned inputs are used for the ANN, and the time taken and efficiency obtained are compared with the previous one. The medical diagnostic data set is taken from UCI medical repository for the hepatitis disease. There is a significant percentage reduction in training time as well as testing time of ANN and a significant improvement in the success rate of diagnosis.
We describe a new approach for face recognition using modular neural networks with a fuzzy logic ... more We describe a new approach for face recognition using modular neural networks with a fuzzy logic method for response integration. We proposed a new architecture for modular neural net-works for achieving pattern recognition in the particular case of human faces. Also, ...
This article investigates a tri-trophic food chain model with the epidemic at the bottom level (p... more This article investigates a tri-trophic food chain model with the epidemic at the bottom level (prey) and gestation delays for the next two levels( the intermediate predator and top predator). The steady states' positivity, boundedness, existence, and stability have been analyzed. The study of Hopf bifurcation has been carried out for different steady states. It is shown that when the gestation delay for intermediate predator τ 1 is absent, the interior equilibrium is locally stable for gestation delays of top predators τ 2 ∈ (0, τ + 20 ) and performs Hopf bifurcation for all τ 2 > τ + 20 . Further, in the presence both the delay τ 1 and τ 2 , the system becomes unstable when τ 1 , τ 2 crosses their critical values τ ++ 10 , τ ++ 20 respectively and executes Hopf bifurcation τ 1 > τ ++ 10 , τ 2 > τ ++ 20 . Moreover, the proposed system is stable for τ 1 > τ ++ 10 , τ 2 < τ ++ 20 and unstable for τ 1 < τ ++ 10 , τ 2 > τ ++ 20 . The sensitivity analysis of the system at coexistence steady state are presented to identify the effect of system parameters on the variables. Finally, we have presented a numerical simulation to verify our analytical findings.
Deep Learning has made great impact on computer vision applications such as object detection, ima... more Deep Learning has made great impact on computer vision applications such as object detection, image classification, segmentation and localization. Computer based system for lung cancer detection and diagnosis helps in detecting the lung nodules and analysing the nodule features for further segmentation and classification. This paper explains the effect of UNet model on various back bone architectures pretrained on ImageNet dataset. This work was carried out on LIDC-IDRI dataset and the efficiency interms of IoU score and accuracy both in training and validation datasets are listed. Results show that EfficientNet backbone outperforms other backbone architectures.
kNN is a very effective Instance based learning method, and it is easy to implement. Due to heter... more kNN is a very effective Instance based learning method, and it is easy to implement. Due to heterogeneous nature of data, noises from different possible sources are also widespread in nature especially in case of large-scale databases. For noise elimination and effect of pseudo neighbours, in this paper, we propose a new learning algorithm which performs the task of anomaly detection and removal of pseudo neighbours from the dataset so as to provide comparative better results. This algorithm also tries to minimize effect of those neighbours which are distant. A concept of certainty measure is also introduced for experimental results. The advantage of using concept of mutual neighbours and distance-weighted voting is that, dataset will be refined after removal of anomaly and weightage concept compels to take into account more consideration of those neighbours, which are closer. Consequently, finally the performance of proposed algorithm is calculated.
Communications in computer and information science, 2020
Image based Lung cancer diagnosis offers an effective and non-intrusive method for identification... more Image based Lung cancer diagnosis offers an effective and non-intrusive method for identification and diagnosis of lung cancer in a patient. This is done by segmentation of lung nodules from radiological images obtained from patients. This paper proposes a framework to segment nodule lesions in images. The novelty of the proposed framework is that we use optical flow information along with the morphological operations on images. The proposed framework applies the Farneback method of optical flow extraction algorithms for the nodule segmentation task. This work validates and demonstrates that optical flow information is useful for nodule segmentation and can be used in tandom with other features.
The Plant Pest problem is a crucial area of investigation for the present time. Myriad chemical s... more The Plant Pest problem is a crucial area of investigation for the present time. Myriad chemical solutions are available to solve this issue, but due to their toxic outcomes, the Natural enemy of pests is come across to be a better solution. This article proposes a fractional calculus approach to a plant pest mature and immature Natural enemy model. Existence and uniqueness of the solution and non-negativity of the solution is being discussed also reached on the Global and Local stability of the equilibrium point. As memory is always present in biological system that is the reason the suggested technique is more advantageous due to the memory effect of fractional derivatives and produces more appropriate solutions. Simulation study for different fractional parameters generates the optimum solution.
In this paper, a non-linear mathematical model with three irresistible classes for the impacts of... more In this paper, a non-linear mathematical model with three irresistible classes for the impacts of media awareness programs on the spread of irresistible infections, for example, influenza, has been proposed and analyzed. In the modeling process, it is expected that illness spreads because of the contact between the susceptibles and infectives, as it were. The growth rate of media awareness programs influencing the populace corresponds to the number of infective people. We examine the dynamical behavior and systematic investigation of the framework for the model, which demonstrates that the model has two equilibrium points, i.e., disease-free equilibrium (DFE) and interior (endemic) equilibrium. The outcomes show that the primary reproduction number determines the dynamics of the model. For the basic reproduction number R 0 < 1, the disease-free equilibrium is locally as well as globally asymptotically stable under a specific parameter set. If R 0 > 1, the model at the interior equilibrium is locally asymptotically stable. At long last, numerical arrangements of the model validate the analytical outcomes and facilitate a sensitivity analysis of the model parameters.
Extremists are increasingly using social media to recruit and radicalize other users and increase... more Extremists are increasingly using social media to recruit and radicalize other users and increase their money. Terrorists can use popular social networks accounts and perform their activities in a hidden way. So, it is crucial to create a fruitful mechanism for controlling the spread of misinformation. Otherwise, a large number of people can mislead by this terrorist activity by joining them. Here, we propose malicious news spreading model incorporating hidden attackers of a social network. A threshold is defined for deciding the extinction of malicious news from a social network. Here, we show the importance of network alertness and activity of cyber-security agencies in the modified model. Moreover, we obtained the optimal values of the control parameters for emergencies.
Zenodo (CERN European Organization for Nuclear Research), Apr 29, 2022
With the emergence of technologies and the increasing number of users of the Internet, such as th... more With the emergence of technologies and the increasing number of users of the Internet, such as the Internet of Things (IoT) paradigm, there are new and modern efforts to invade networks and computer systems. Many researchers have developed various artificial intelligence-based algorithms to detect these attacks. For network security they focus on machine learning models that are used in IoT and intrusion detection. In this study, we have proposed a machine learning based intrusion detection system which is a combination of support vector machine (SVM) and Genetic Algorithm (GA). GA is basically used for feature selection and parameter optimization of SVM models. The performance of a GA-based SVM model is evaluated on an KDD Cup 99 intrusion database. First, GA optimizes the KDD Cup 99 dataset and then optimizes the weight and parameters of the SVM model. The experimental result presents that the SVM model gives prominence in terms of detection rate, accuracy, false positive rate and false negative rate and also compared to other literature works.
Seasonal variability strongly affects the animal population in wildlife. It becomes essential to ... more Seasonal variability strongly affects the animal population in wildlife. It becomes essential to model seasonality in eco-epidemic dynamics to know the effect of system parameters in a periodic environment. This article presents a set of non-autonomous differential equations with time-varying disease transmission rates among prey and predators, the mortality rate of a diseased predator, the predation rate of healthy prey, and an additional food supply. The positiveness, boundedness, and presence of solution are derived. We have proved that the infection-free state is stable if periodic basic reproduction number $$R_C(t)<1$$ R C ( t ) < 1 . The stability of the coexistence state is shown at $$R_C(t)>1$$ R C ( t ) > 1 using the Poincare map and comparison theory. The significance of the parameters related to disease transmission and prevalence is described using sensitivity analysis. Numerical simulation verified our analytical findings and proved that the predator control...
The pandemic being a health issue at its core is a multifaceted crisis encompassing both economic... more The pandemic being a health issue at its core is a multifaceted crisis encompassing both economic and epidemic factors in a twisted tale of challenges. In counteraction, we have proposed a combined epidemic-economic model that analyses system dynamics arising in the presence of an infectious disease (SARS-2-COVID-19 in our case). Dynamical analysis of the system has been performed in context to the equilibria along with local and global stability analysis of the system simultaneously visualizing the effect on capital stabilization. The global stability analysis has been performed using graph-theoretic method. Curve-Fit has been performed for the system using optimization algorithm. The relation between all the parameters and variables involved in the model has been explored by calculating sensitivity indices which gives us the proportion that a relative change in a parameter brings to the relative change in a variable. Our findings reveal that (1) Vaccination instigates economic growth (with evidence of data obtained for 24 countries). ( ) Complete vaccination leads to a considerable reduction in all infections (reduction up to 90%, as per current CDC study). ( ) Excessive exposure to media can facilitate spike in infections. (4) Parameter sensitivity analysis can be of immense help in policy formation.
Proceedings of the 2018 International Conference on Algorithms, Computing and Artificial Intelligence, 2018
In this paper, we present a deep learning based approach to assign POS tags to words in a piece o... more In this paper, we present a deep learning based approach to assign POS tags to words in a piece of text given to it as input. We propose an unsupervised approach owing to the lack of a large Sanskrit annotated corpora and use the untagged Sanskrit Corpus prepared by JNU for our purpose. The only tagged corpora for Sanskrit is created by JNU which has 115,000 words which are not sufficient to apply supervised deep learning approaches. For the tag assignment purpose and determining model accuracy, we utilize this tagged corpus. We explore various methods through which each Sanskrit word can be represented as a point multi-dimensional vector space whose position accurately captures its meaning and semantic information associated with it. We also explore other data sources to improve performance and robustness of the vector representations. We use these rich vector representations and explore autoencoder based approaches for dimensionality reduction to compress these into encodings which are suitable for clustering in the vector space. We experiment with different dimensions of these compressed representations and present one which was found to offer the best clustering performance. For modelling the sequence in order to preserve the semantic information we feed these embeddings to a bidirectional LSTM autoencoder. We assign a POS tag to each of the obtained clusters and produce our result by testing the model on the tagged corpus.
kNN is a very effective Instance based learning method, and it is easy to implement. Due to heter... more kNN is a very effective Instance based learning method, and it is easy to implement. Due to heterogeneous nature of data, noises from different possible sources are also widespread in nature especially in case of large-scale databases. For noise elimination and effect of pseudo neighbours, in this paper, we propose a new learning algorithm which performs the task of anomaly detection and removal of pseudo neighbours from the dataset so as to provide comparative better results. This algorithm also tries to minimize effect of those neighbours which are distant. A concept of certainty measure is also introduced for experimental results. The advantage of using concept of mutual neighbours and distance-weighted voting is that, dataset will be refined after removal of anomaly and weightage concept compels to take into account more consideration of those neighbours, which are closer. Consequently, finally the performance of proposed algorithm is calculated.
Uploads
Papers by Joydip Dhar