Skip to main content

Suresh Dara

DIT University Dehradun, Computer Science, Faculty Member

Indian School of Mines, Dhanbad, Computer Science and Engineering, Graduate Student

Followers

43

Following

21

Co-author

1

Public Views

Phone: 9410141469
Address: Associate Professor/CSE
DIT University

less

Lebanese University

Nicolas Lanchier

Universidad Viña del Mar

Florentin Smarandache

University of New Mexico

Camino Balbuena

Universidad Politecnica de Catalunya

Darine Al Mniny

Rajshahi University of Engineering and Technology

Interests

Uploads

Papers by Suresh Dara

Extending some results on the second neighborhood conjecture

arXiv: Discrete Mathematics, 2018

A vertex in a directed graph is said to have a large second neighborhood if it has at least as ma... more A vertex in a directed graph is said to have a large second neighborhood if it has at least as many second out-neighbors as out-neighbors. The Second Neighborhood Conjecture, first stated by Seymour, asserts that there is a vertex having a large second neighborhood in every oriented graph (a directed graph without loops or digons). We prove that oriented graphs whose missing edges can be partitioned into a (possibly empty) matching and a (possibly empty) star satisfy this conjecture. This generalizes a result of Fidler and Yuster. An implication of our result is that every oriented graph without a sink and whose missing edges form a (possibly empty) matching has at least two vertices with large second neighborhoods. This is a strengthening of a theorem of Havet and Thomasse, who showed that the same holds for tournaments without a sink. Moreover, we also show that the conjecture is true for oriented graphs whose vertex set can be partitioned into an independent set and a 2-degenerat...

Smart Heart Disease Prediction System with IoT and Fog Computing Sectors Enabled by Cascaded Deep Learning Model

Computational Intelligence and Neuroscience

Chronic illnesses like chronic respiratory disease, cancer, heart disease, and diabetes are threa... more Chronic illnesses like chronic respiratory disease, cancer, heart disease, and diabetes are threats to humans around the world. Among them, heart disease with disparate features or symptoms complicates diagnosis. Because of the emergence of smart wearable gadgets, fog computing and “Internet of Things” (IoT) solutions have become necessary for diagnosis. The proposed model integrates Edge-Fog-Cloud computing for the accurate and fast delivery of outcomes. The hardware components collect data from different patients. The heart feature extraction from signals is done to get significant features. Furthermore, the feature extraction of other attributes is also gathered. All these features are gathered and subjected to the diagnostic system using an Optimized Cascaded Convolution Neural Network (CCNN). Here, the hyperparameters of CCNN are optimized by the Galactic Swarm Optimization (GSO). Through the performance analysis, the precision of the suggested GSO-CCNN is 3.7%, 3.7%, 3.6%, 7.6...

Hamming Distance based Binary PSO for Feature Selection and Classification from high dimensional Gene Expression Data

In this article, a Binary Particle Swarm Optimization (BPSO) algorithm is proposed incorporating ... more In this article, a Binary Particle Swarm Optimization (BPSO) algorithm is proposed incorporating hamming distance as a distance measure between particles for feature selection problem from high di- mensional microarray gene expression data. Hamming distance is used as an similarity measurement for updating the velocities of each par- ticles or solutions. It also helps to reduce extra parameter (i.e. Vmin) as needed in conventional BPSO during velocity updation. An initial fast pre-processing heuristic method is used for crude domain reduction from high dimension. Then the tness function is suitably designed in multi objective framework for further reduction and soft tuning on the reduced features using BPSO. The performance of the proposed method is tested on three benchmark cancerous datasets (i.e., colon, lymphoma and leukemia cancer). The comparative study is also performed on the existing literature to show the eectiveness of the proposed method.

Evolutionary Computation based Feature Selection: A Survey

2018 Second International Conference on Electronics, Communication and Aerospace Technology (ICECA), 2018

In previous years, different Lateral thinking optimization techniques have been developed based o... more In previous years, different Lateral thinking optimization techniques have been developed based on evolutionary computation. Many of these methods are inspired by spill out behaviors in nature. In this Paper, a new optimization algorithm based on the law of gravity and mass interactions named as Gravitational Search Algorithm (GSA) is discussed for solving feature selection. In GSA, the searcher agents are a collection of masses which will interact with each other based on the law of motion and Newtonian gravity which gives the binary evolutionary optimized high performance. The detailed feature selection has been discussed in this paper and The GSA method has been compared with some well-known optimized search methods such as GA (Genetic Algorithm), PSO(Particle Swarm Optimization).

Artificial Bee Colony Algorithm : A Survey and Recent Applications

Swarm intelligence has provided solution to classical mathematical techniques which can’t be deal... more Swarm intelligence has provided solution to classical mathematical techniques which can’t be dealt easily. An intelligent foraging behavior of honey bees falls into category of swarm intelligence. Deception of honey bee foraging behavior of Artificial Bee Colony(ABC) was established in 2005. Due to its origination, researches had been made on ABC and applied it on different types of problems. In this article, a brief review is presenting about ABC and its progress, implementation, comparative completion and future scrutinizing perspectives. We focused and presented on recent works by ABC and It’s applications.

Clustering-based hybrid feature selection approach for high dimensional microarray data

Chemometrics and Intelligent Laboratory Systems

Machine Learning in Drug Discovery: A Review

Artificial Intelligence Review

This review provides the feasible literature on drug discovery through ML tools and techniques th... more This review provides the feasible literature on drug discovery through ML tools and techniques that are enforced in every phase of drug development to accelerate the research process and deduce the risk and expenditure in clinical trials. Machine learning techniques improve the decision-making in pharmaceutical data across various applications like QSAR analysis, hit discoveries, de novo drug architectures to retrieve accurate outcomes. Target validation, prognostic biomarkers, digital pathology are considered under problem statements in this review. ML challenges must be applicable for the main cause of inadequacy in interpretability outcomes that may restrict the applications in drug discovery. In clinical trials, absolute and methodological data must be generated to tackle many puzzles in validating ML techniques, improving decision-making, promoting awareness in ML approaches, and deducing risk failures in drug discovery.

Cancer data classification by quantum-inspired immune clone optimization-based optimal feature selection using gene expression data: deep learning approach

Data Technologies and Applications

PurposeGene selection is considered as the fundamental process in the bioinformatics field. The e... more PurposeGene selection is considered as the fundamental process in the bioinformatics field. The existing methodologies pertain to cancer classification are mostly clinical basis, and its diagnosis capability is limited. Nowadays, the significant problems of cancer diagnosis are solved by the utilization of gene expression data. The researchers have been introducing many possibilities to diagnose cancer appropriately and effectively. This paper aims to develop the cancer data classification using gene expression data.Design/methodology/approachThe proposed classification model involves three main phases: “(1) Feature extraction, (2) Optimal Feature Selection and (3) Classification”. Initially, five benchmark gene expression datasets are collected. From the collected gene expression data, the feature extraction is performed. To diminish the length of the feature vectors, optimal feature selection is performed, for which a new meta-heuristic algorithm termed as quantum-inspired immune ...

Feature Extraction By Using Deep Learning: A Survey

2018 Second International Conference on Electronics, Communication and Aerospace Technology (ICECA)

Deep learning is presently an effective research area in machine learning technique and pattern c... more Deep learning is presently an effective research area in machine learning technique and pattern classification association. This has achieved big success in the areas of application namely computer vision, speech recognition, and NLP. This paper gives the impact of feature extraction that used in a deep learning technique such as Convolutional Neural Network (CNN). The purpose of this paper presents an emerged survey of actual literatures on feature extraction methods since past five years. As the raising of application demand increases, a large study and analysis in the feature extraction field became very efficient. In this paper, we presented a detailed study on deep learning, and a described some of existing methodology of feature extraction.

An Efficient Algorithm for Feature Selection Problem in Gene Expression Data: A Spider Monkey Optimization Approach

SSRN Electronic Journal

In this paper, we presented an adequate algorithm in gene expression data for feature selection. ... more In this paper, we presented an adequate algorithm in gene expression data for feature selection. A Spider Monkey Optimization (SMO) based approach has implemented to get feature selection subset from gene expression data of high dimension. In consideration of high dimensionality data where lie large amount of superfluous features, we are using a well defined fast preprocessing heuristic based approach for diminishing features. Fitness function can be illustrated as it supports SMO scheme to knob the inconsistent targets i.e., to decrease feature cardinality and preserving unique efficiency (i.e., accuracy of classification). The bench mark datasets of microarray gene expression have used and demonstrated the results along in-depth provisional research using a classifier named k-NN.

Efficient Algorithms for Hotspot Problem in Wireless Sensor Networks: Gravitational Search Algorithm

Advances in Intelligent Systems and Computing

Energy conservation of sensor nodes (SNs) is the major concern of wireless sensor networks (WSNs)... more Energy conservation of sensor nodes (SNs) is the major concern of wireless sensor networks (WSNs) as those are operated by small batteries with a limited power. In a clustered WSN, cluster heads (CHs) collects local information such as temperature, humidity, pressure etc. from the member SNs aggregate it and send to the sink through few intermediate CHs. Here, the CHs that are closer to the sink are over burdened as they are responsible for forwarding more number of packets than the farther CHs that tends to exhaust their energy quickly. This results in network partitioning and this problem well known hot spot or energy hole problem. In this paper, a Gravitational Search Algorithm (GSA) approach based clustering and routing algorithms are proposed to address the hot spot problem. In clustering, we select few efficient SNs as CHs from the normal SNs with respect to certain cost function. We design an algorithm for CH selection based on GSA and assign the remaining SNs to the CHs based on another derived cost function. Then, a GSA based routing algorithm is presented with respect to the routing cost function. These algorithms are intended to develop to enhance the lifetime of network with efficient encoding schemes of GSA. The proposed algorithms are simulated on various scenarios of WSNs by varying number of SNs. The results of the proposed algorithms are compared with few well known algorithms to show the supremacy in terms network lifetime, residual energy and number of alive SNs.

Cancer microarray data feature selection using multi-objective binary particle swarm optimization algorithm

EXCLI journal, 2016

Cancer investigations in microarray data play a major role in cancer analysis and the treatment. ... more Cancer investigations in microarray data play a major role in cancer analysis and the treatment. Cancer microarray data consists of complex gene expressed patterns of cancer. In this article, a Multi-Objective Binary Particle Swarm Optimization (MOBPSO) algorithm is proposed for analyzing cancer gene expression data. Due to its high dimensionality, a fast heuristic based pre-processing technique is employed to reduce some of the crude domain features from the initial feature set. Since these pre-processed and reduced features are still high dimensional, the proposed MOBPSO algorithm is used for finding further feature subsets. The objective functions are suitably modeled by optimizing two conflicting objectives i.e., cardinality of feature subsets and distinctive capability of those selected subsets. As these two objective functions are conflicting in nature, they are more suitable for multi-objective modeling. The experiments are carried out on benchmark gene expression datasets, i...

A Rough Based Hybrid Binary PSO Algorithm for Flat Feature Selection and Classification in Gene Expression Data

Annals of Data Science, 2017

Feature selection in high dimensional data, particularly, in gene expression data, is one of the ... more Feature selection in high dimensional data, particularly, in gene expression data, is one of the challenging task in bioinformatics due to the curse of dimensionality, data redundancy and noise values. In gene expression data, insignificant features causes poor classification, hence feature selection reduces feature subset, improving classification accuracy. Feature selection algorithms in gene expression data(such as filter based, wrapper based and hybrid methods) performing poor accuracy, where as few methods takes too much time to converge for an acceptable results. For example, in NSGA-II, over 10,000 generations, on an average, to converge in the search space. where it incurs increased computational time. Proposed rough based hybrid binary PSO algorithm, which uses a heuristic based fast processing strategy to reduce crude domain features by statistical elimination of redundant features and then discretized subsequently into a binary table, known as distinction table, in rough set theory. This distinction table is later used as input to evaluate and optimize the objectives functions i.e., to generate reduct in rough set theory. The proposed hybrid binary PSO is then used to tune the objective functions, to choose the most important features (i:e:reduct). The fitness function is used in such a way that it can reduce the cardinality of the features and at the same time, improve the classification performance as well. Results have been demonstrated to show the effectiveness of the proposed method, on existing three benchmark datasets (i.e. colon cancer, lymphoma and leukemia data), from literature.

An Improved Quantum Inspired Immune Clone Optimization Algorithm

Lecture Notes in Computer Science, 2016

An improved quantum inspired immune clone optimization algorithm is proposed for optimization pro... more An improved quantum inspired immune clone optimization algorithm is proposed for optimization problem. It is proposed based on the immune clone algorithm and quantum computing theory. The algorithm adopts the quantum bit to express the chromosomes, and uses the quantum gate updating to implement evolutionary of population which can take advantage of the parallelism of quantum computing and the learning, memory capability of the immune system. Quantum observing entropy is introduced to evaluate the population evolutionary level, and relevant parameters are adjusted according to the entropy value. The proposed algorithm is tested on few benchmark optimization functions and the results are compared with other existing algorithms. The simulation results show that the proposed algorithm has better convergence, robustness and precision.

Feature Extraction In Medical Images by Using Deep Learning Approach

Deep learning is presently an effective research area in machine learning technique and pattern c... more Deep learning is presently an effective research area in machine learning technique and pattern classification association. This has achieved big success in the areas of application namely computer vision, speech recognition, and NLP. This paper gives the impact of feature extraction that used in a deep learning technique such as Convolutional Neural Network (CNN). In this paper, feature extraction method is proposed and performed on medical images which CT scan Cancer datasetss. The experimental results have presented with proposed approach.

An Elitist Binary PSO Algorithm for Selecting Features in High Dimensional Data

by Suresh Dara and Haider Banka

Smart Innovation, Systems and Technologies, 2014

A binary PSO feature selection algorithm for gene expression data

by Suresh Dara and Haider Banka

2014 International Conference on Advances in Communication and Computing Technologies (ICACACT 2014), 2014

Feature Selection and Classification for Gene Expression Data using Evolutionary Computation

An evolutionary rough feature selection algorithm is proposed for classifying gene expression pat... more An evolutionary rough feature selection algorithm is
proposed for classifying gene expression patterns. Since the data typically consist of a large number of redundant features, an initial redundancy reduction of the attributes is done to enable faster convergence. Rough set theory is employed to generate the distinction table that enable PSO to find reducts, which represent the minimal sets of non-redundant features capable of discerning between all objects. The effectiveness of the algorithm is demonstrated on three benchmark cancer datasets viz. Colon, Lymphoma and Leukemia using MOGA.

Hamming Distance based Binary PSO for Feature Selection and Classication from high dimensional Gene Expression Data

In this article, a Binary Particle Swarm Optimization (BPSO) algorithm is proposed incorporating ... more In this article, a Binary Particle Swarm Optimization (BPSO)
algorithm is proposed incorporating hamming distance as a distance measure between particles for feature selection problem from high dimensional microarray gene expression data. Hamming distance is used
as an similarity measurement for updating the velocities of each par-
ticles or solutions. It also helps to reduce extra parameter (i.e. Vmin)
as needed in conventional BPSO during velocity updation. An initial fast pre-processing heuristic method is used for crude domain reduction from high dimension. Then the tness function is suitably designed in multi objective framework for further reduction and soft tuning on the reduced features using BPSO. The performance of the proposed method
is tested on three benchmark cancerous datasets (i.e., colon, lymphoma and leukemia cancer). The comparative study is also performed on the existing literature to show the eectiveness of the proposed method.

Extending some results on the second neighborhood conjecture

arXiv: Discrete Mathematics, 2018

A vertex in a directed graph is said to have a large second neighborhood if it has at least as ma... more A vertex in a directed graph is said to have a large second neighborhood if it has at least as many second out-neighbors as out-neighbors. The Second Neighborhood Conjecture, first stated by Seymour, asserts that there is a vertex having a large second neighborhood in every oriented graph (a directed graph without loops or digons). We prove that oriented graphs whose missing edges can be partitioned into a (possibly empty) matching and a (possibly empty) star satisfy this conjecture. This generalizes a result of Fidler and Yuster. An implication of our result is that every oriented graph without a sink and whose missing edges form a (possibly empty) matching has at least two vertices with large second neighborhoods. This is a strengthening of a theorem of Havet and Thomasse, who showed that the same holds for tournaments without a sink. Moreover, we also show that the conjecture is true for oriented graphs whose vertex set can be partitioned into an independent set and a 2-degenerat...

Smart Heart Disease Prediction System with IoT and Fog Computing Sectors Enabled by Cascaded Deep Learning Model

Computational Intelligence and Neuroscience

Chronic illnesses like chronic respiratory disease, cancer, heart disease, and diabetes are threa... more Chronic illnesses like chronic respiratory disease, cancer, heart disease, and diabetes are threats to humans around the world. Among them, heart disease with disparate features or symptoms complicates diagnosis. Because of the emergence of smart wearable gadgets, fog computing and “Internet of Things” (IoT) solutions have become necessary for diagnosis. The proposed model integrates Edge-Fog-Cloud computing for the accurate and fast delivery of outcomes. The hardware components collect data from different patients. The heart feature extraction from signals is done to get significant features. Furthermore, the feature extraction of other attributes is also gathered. All these features are gathered and subjected to the diagnostic system using an Optimized Cascaded Convolution Neural Network (CCNN). Here, the hyperparameters of CCNN are optimized by the Galactic Swarm Optimization (GSO). Through the performance analysis, the precision of the suggested GSO-CCNN is 3.7%, 3.7%, 3.6%, 7.6...

Hamming Distance based Binary PSO for Feature Selection and Classification from high dimensional Gene Expression Data

In this article, a Binary Particle Swarm Optimization (BPSO) algorithm is proposed incorporating ... more In this article, a Binary Particle Swarm Optimization (BPSO) algorithm is proposed incorporating hamming distance as a distance measure between particles for feature selection problem from high di- mensional microarray gene expression data. Hamming distance is used as an similarity measurement for updating the velocities of each par- ticles or solutions. It also helps to reduce extra parameter (i.e. Vmin) as needed in conventional BPSO during velocity updation. An initial fast pre-processing heuristic method is used for crude domain reduction from high dimension. Then the tness function is suitably designed in multi objective framework for further reduction and soft tuning on the reduced features using BPSO. The performance of the proposed method is tested on three benchmark cancerous datasets (i.e., colon, lymphoma and leukemia cancer). The comparative study is also performed on the existing literature to show the eectiveness of the proposed method.

Evolutionary Computation based Feature Selection: A Survey

2018 Second International Conference on Electronics, Communication and Aerospace Technology (ICECA), 2018

In previous years, different Lateral thinking optimization techniques have been developed based o... more In previous years, different Lateral thinking optimization techniques have been developed based on evolutionary computation. Many of these methods are inspired by spill out behaviors in nature. In this Paper, a new optimization algorithm based on the law of gravity and mass interactions named as Gravitational Search Algorithm (GSA) is discussed for solving feature selection. In GSA, the searcher agents are a collection of masses which will interact with each other based on the law of motion and Newtonian gravity which gives the binary evolutionary optimized high performance. The detailed feature selection has been discussed in this paper and The GSA method has been compared with some well-known optimized search methods such as GA (Genetic Algorithm), PSO(Particle Swarm Optimization).

Artificial Bee Colony Algorithm : A Survey and Recent Applications

Swarm intelligence has provided solution to classical mathematical techniques which can’t be deal... more Swarm intelligence has provided solution to classical mathematical techniques which can’t be dealt easily. An intelligent foraging behavior of honey bees falls into category of swarm intelligence. Deception of honey bee foraging behavior of Artificial Bee Colony(ABC) was established in 2005. Due to its origination, researches had been made on ABC and applied it on different types of problems. In this article, a brief review is presenting about ABC and its progress, implementation, comparative completion and future scrutinizing perspectives. We focused and presented on recent works by ABC and It’s applications.

Clustering-based hybrid feature selection approach for high dimensional microarray data

Chemometrics and Intelligent Laboratory Systems

Machine Learning in Drug Discovery: A Review

Artificial Intelligence Review

This review provides the feasible literature on drug discovery through ML tools and techniques th... more This review provides the feasible literature on drug discovery through ML tools and techniques that are enforced in every phase of drug development to accelerate the research process and deduce the risk and expenditure in clinical trials. Machine learning techniques improve the decision-making in pharmaceutical data across various applications like QSAR analysis, hit discoveries, de novo drug architectures to retrieve accurate outcomes. Target validation, prognostic biomarkers, digital pathology are considered under problem statements in this review. ML challenges must be applicable for the main cause of inadequacy in interpretability outcomes that may restrict the applications in drug discovery. In clinical trials, absolute and methodological data must be generated to tackle many puzzles in validating ML techniques, improving decision-making, promoting awareness in ML approaches, and deducing risk failures in drug discovery.

Cancer data classification by quantum-inspired immune clone optimization-based optimal feature selection using gene expression data: deep learning approach

Data Technologies and Applications

PurposeGene selection is considered as the fundamental process in the bioinformatics field. The e... more PurposeGene selection is considered as the fundamental process in the bioinformatics field. The existing methodologies pertain to cancer classification are mostly clinical basis, and its diagnosis capability is limited. Nowadays, the significant problems of cancer diagnosis are solved by the utilization of gene expression data. The researchers have been introducing many possibilities to diagnose cancer appropriately and effectively. This paper aims to develop the cancer data classification using gene expression data.Design/methodology/approachThe proposed classification model involves three main phases: “(1) Feature extraction, (2) Optimal Feature Selection and (3) Classification”. Initially, five benchmark gene expression datasets are collected. From the collected gene expression data, the feature extraction is performed. To diminish the length of the feature vectors, optimal feature selection is performed, for which a new meta-heuristic algorithm termed as quantum-inspired immune ...

Feature Extraction By Using Deep Learning: A Survey

2018 Second International Conference on Electronics, Communication and Aerospace Technology (ICECA)

Deep learning is presently an effective research area in machine learning technique and pattern c... more Deep learning is presently an effective research area in machine learning technique and pattern classification association. This has achieved big success in the areas of application namely computer vision, speech recognition, and NLP. This paper gives the impact of feature extraction that used in a deep learning technique such as Convolutional Neural Network (CNN). The purpose of this paper presents an emerged survey of actual literatures on feature extraction methods since past five years. As the raising of application demand increases, a large study and analysis in the feature extraction field became very efficient. In this paper, we presented a detailed study on deep learning, and a described some of existing methodology of feature extraction.

An Efficient Algorithm for Feature Selection Problem in Gene Expression Data: A Spider Monkey Optimization Approach

SSRN Electronic Journal

In this paper, we presented an adequate algorithm in gene expression data for feature selection. ... more In this paper, we presented an adequate algorithm in gene expression data for feature selection. A Spider Monkey Optimization (SMO) based approach has implemented to get feature selection subset from gene expression data of high dimension. In consideration of high dimensionality data where lie large amount of superfluous features, we are using a well defined fast preprocessing heuristic based approach for diminishing features. Fitness function can be illustrated as it supports SMO scheme to knob the inconsistent targets i.e., to decrease feature cardinality and preserving unique efficiency (i.e., accuracy of classification). The bench mark datasets of microarray gene expression have used and demonstrated the results along in-depth provisional research using a classifier named k-NN.

Efficient Algorithms for Hotspot Problem in Wireless Sensor Networks: Gravitational Search Algorithm

Advances in Intelligent Systems and Computing

Energy conservation of sensor nodes (SNs) is the major concern of wireless sensor networks (WSNs)... more Energy conservation of sensor nodes (SNs) is the major concern of wireless sensor networks (WSNs) as those are operated by small batteries with a limited power. In a clustered WSN, cluster heads (CHs) collects local information such as temperature, humidity, pressure etc. from the member SNs aggregate it and send to the sink through few intermediate CHs. Here, the CHs that are closer to the sink are over burdened as they are responsible for forwarding more number of packets than the farther CHs that tends to exhaust their energy quickly. This results in network partitioning and this problem well known hot spot or energy hole problem. In this paper, a Gravitational Search Algorithm (GSA) approach based clustering and routing algorithms are proposed to address the hot spot problem. In clustering, we select few efficient SNs as CHs from the normal SNs with respect to certain cost function. We design an algorithm for CH selection based on GSA and assign the remaining SNs to the CHs based on another derived cost function. Then, a GSA based routing algorithm is presented with respect to the routing cost function. These algorithms are intended to develop to enhance the lifetime of network with efficient encoding schemes of GSA. The proposed algorithms are simulated on various scenarios of WSNs by varying number of SNs. The results of the proposed algorithms are compared with few well known algorithms to show the supremacy in terms network lifetime, residual energy and number of alive SNs.

Cancer microarray data feature selection using multi-objective binary particle swarm optimization algorithm

EXCLI journal, 2016

Cancer investigations in microarray data play a major role in cancer analysis and the treatment. ... more Cancer investigations in microarray data play a major role in cancer analysis and the treatment. Cancer microarray data consists of complex gene expressed patterns of cancer. In this article, a Multi-Objective Binary Particle Swarm Optimization (MOBPSO) algorithm is proposed for analyzing cancer gene expression data. Due to its high dimensionality, a fast heuristic based pre-processing technique is employed to reduce some of the crude domain features from the initial feature set. Since these pre-processed and reduced features are still high dimensional, the proposed MOBPSO algorithm is used for finding further feature subsets. The objective functions are suitably modeled by optimizing two conflicting objectives i.e., cardinality of feature subsets and distinctive capability of those selected subsets. As these two objective functions are conflicting in nature, they are more suitable for multi-objective modeling. The experiments are carried out on benchmark gene expression datasets, i...

A Rough Based Hybrid Binary PSO Algorithm for Flat Feature Selection and Classification in Gene Expression Data

Annals of Data Science, 2017

Feature selection in high dimensional data, particularly, in gene expression data, is one of the ... more Feature selection in high dimensional data, particularly, in gene expression data, is one of the challenging task in bioinformatics due to the curse of dimensionality, data redundancy and noise values. In gene expression data, insignificant features causes poor classification, hence feature selection reduces feature subset, improving classification accuracy. Feature selection algorithms in gene expression data(such as filter based, wrapper based and hybrid methods) performing poor accuracy, where as few methods takes too much time to converge for an acceptable results. For example, in NSGA-II, over 10,000 generations, on an average, to converge in the search space. where it incurs increased computational time. Proposed rough based hybrid binary PSO algorithm, which uses a heuristic based fast processing strategy to reduce crude domain features by statistical elimination of redundant features and then discretized subsequently into a binary table, known as distinction table, in rough set theory. This distinction table is later used as input to evaluate and optimize the objectives functions i.e., to generate reduct in rough set theory. The proposed hybrid binary PSO is then used to tune the objective functions, to choose the most important features (i:e:reduct). The fitness function is used in such a way that it can reduce the cardinality of the features and at the same time, improve the classification performance as well. Results have been demonstrated to show the effectiveness of the proposed method, on existing three benchmark datasets (i.e. colon cancer, lymphoma and leukemia data), from literature.

An Improved Quantum Inspired Immune Clone Optimization Algorithm

Lecture Notes in Computer Science, 2016

An improved quantum inspired immune clone optimization algorithm is proposed for optimization pro... more An improved quantum inspired immune clone optimization algorithm is proposed for optimization problem. It is proposed based on the immune clone algorithm and quantum computing theory. The algorithm adopts the quantum bit to express the chromosomes, and uses the quantum gate updating to implement evolutionary of population which can take advantage of the parallelism of quantum computing and the learning, memory capability of the immune system. Quantum observing entropy is introduced to evaluate the population evolutionary level, and relevant parameters are adjusted according to the entropy value. The proposed algorithm is tested on few benchmark optimization functions and the results are compared with other existing algorithms. The simulation results show that the proposed algorithm has better convergence, robustness and precision.

Feature Extraction In Medical Images by Using Deep Learning Approach

Deep learning is presently an effective research area in machine learning technique and pattern c... more Deep learning is presently an effective research area in machine learning technique and pattern classification association. This has achieved big success in the areas of application namely computer vision, speech recognition, and NLP. This paper gives the impact of feature extraction that used in a deep learning technique such as Convolutional Neural Network (CNN). In this paper, feature extraction method is proposed and performed on medical images which CT scan Cancer datasetss. The experimental results have presented with proposed approach.

An Elitist Binary PSO Algorithm for Selecting Features in High Dimensional Data

by Suresh Dara and Haider Banka

Smart Innovation, Systems and Technologies, 2014

A binary PSO feature selection algorithm for gene expression data

by Suresh Dara and Haider Banka

2014 International Conference on Advances in Communication and Computing Technologies (ICACACT 2014), 2014

Feature Selection and Classification for Gene Expression Data using Evolutionary Computation

An evolutionary rough feature selection algorithm is proposed for classifying gene expression pat... more An evolutionary rough feature selection algorithm is
proposed for classifying gene expression patterns. Since the data typically consist of a large number of redundant features, an initial redundancy reduction of the attributes is done to enable faster convergence. Rough set theory is employed to generate the distinction table that enable PSO to find reducts, which represent the minimal sets of non-redundant features capable of discerning between all objects. The effectiveness of the algorithm is demonstrated on three benchmark cancer datasets viz. Colon, Lymphoma and Leukemia using MOGA.

Hamming Distance based Binary PSO for Feature Selection and Classication from high dimensional Gene Expression Data

In this article, a Binary Particle Swarm Optimization (BPSO) algorithm is proposed incorporating ... more In this article, a Binary Particle Swarm Optimization (BPSO)
algorithm is proposed incorporating hamming distance as a distance measure between particles for feature selection problem from high dimensional microarray gene expression data. Hamming distance is used
as an similarity measurement for updating the velocities of each par-
ticles or solutions. It also helps to reduce extra parameter (i.e. Vmin)
as needed in conventional BPSO during velocity updation. An initial fast pre-processing heuristic method is used for crude domain reduction from high dimension. Then the tness function is suitably designed in multi objective framework for further reduction and soft tuning on the reduced features using BPSO. The performance of the proposed method
is tested on three benchmark cancerous datasets (i.e., colon, lymphoma and leukemia cancer). The comparative study is also performed on the existing literature to show the eectiveness of the proposed method.