2010 Ninth IEEE International Symposium on Network Computing and Applications, 2010
Existing data management schemes in clouds are mainly based on Google File System (GFS) and MapRe... more Existing data management schemes in clouds are mainly based on Google File System (GFS) and MapReduce. Problems arise when data partitioning among numerous available nodes therein. This research paper explores new methods of partitioning and distributing data, that is, resource virtualization in cloud computing. Loosely-coupled associative computing techniques, which have so far not been considered for clouds, can provide the
2011 11th International Conference on Intelligent Systems Design and Applications, 2011
Abstract Network intrusion detection has remained a field of rigorous research over the past few ... more Abstract Network intrusion detection has remained a field of rigorous research over the past few years. Advances in computing performance, in terms of processing power and storage, have allowed the use of resource-intensive intelligent algorithms, to detect intrusive ...
2011 11th International Conference on Hybrid Intelligent Systems (HIS), 2011
Multiple-feature implementation enables a holistic approach towards pattern recognition procedure... more Multiple-feature implementation enables a holistic approach towards pattern recognition procedure that takes into consideration all significant features, which represent a particular set of complex patterns, such as images and sensor readings. This intends to reduce the bias effect of selecting only a single feature for classification/recognition purposes. In this paper we demonstrate the effectiveness of this approach in comparison with
2012 12th International Conference on Control Automation Robotics & Vision (ICARCV), 2012
ABSTRACT In Wireless Sensor Networks (WSNs), using physically sensed data for accurate automated ... more ABSTRACT In Wireless Sensor Networks (WSNs), using physically sensed data for accurate automated decision making is challenging. In response to these challenges, a combined Genetic Algorithm (GA) and pattern recognition scheme (PR) is presented in this paper. The aim of the scheme is to reduce the exponential relationship between problem size and time complexity of GA for guiding robots using WSN. The PR scheme presented in this paper is called Cellular Weighted Pattern Recogniser (CWPR) that simplifies computations and communications for energy conservation and speeds up recognition by leveraging the parallel distributed processing capabilities of WSN. Additionally, CWPR solves the problem of dilation, translation, and rotation to provide efficient pattern recognition in energy constrained WSN environments. Combining CWPR with GA allows GA to learn from experience and solve similar problems in fewer number of generations. The experimental results show that the approach efficiently supports a variety of PR applications for WSN guided robots.
2008 Ninth International Conference on Parallel and Distributed Computing, Applications and Technologies, 2008
... Wireless Sensor Networks Anang Hudaya Muhamad Amin and Asad I. Khan Clayton School of Informa... more ... Wireless Sensor Networks Anang Hudaya Muhamad Amin and Asad I. Khan Clayton School of Information Technology Monash University Clayton, 3168 VIC Australia {Anang.Hudaya, Asad.Khan}@infotech.monash.edu.au Abstract ...
2012 International Conference on Computer & Information Science (ICCIS), 2012
ABSTRACT Reduced-Distributed Hierarchical Graph Neuron (R-DHGN) is a one-shot learning distribute... more ABSTRACT Reduced-Distributed Hierarchical Graph Neuron (R-DHGN) is a one-shot learning distributed associative memory algorithm for data classification, which reduces the computational complexity of existing recognition algorithms by distributing the recognition process into smaller processing clusters. This paper investigates an effect of unsupervised one-shot learning mechanism for data classification within a computational network. This computational network may represent a network of objects than can be deployed in the existing Internet-of-Things (IoT) environment that offers seamless connectivity between smart devices such as sensors. Our approach extends the pattern recognition capability of Distributed Hierarchical Graph Neuron (DHGN). The interprocess communications of DHGN scheme is significantly reduced, and preliminary results obtained from the series of comparative analyses with other established classifiers have indicated the capability of R-DHGN to produce one-shot classification technique using a lightweight recognition mechanism. Simple dataset of iris plants have been used to demonstrate such capability of R-DHGN.
Page 401. Single-Cycle Image Recognition Using an Adaptive Granularity Associative Memory Network... more Page 401. Single-Cycle Image Recognition Using an Adaptive Granularity Associative Memory Network Anang Hudaya Muhamad Amin and Asad I. Khan Clayton School of Information Technology, Monash University, Clayton, 3168 VIC, Australia {Anang. Hudaya, Asad. ...
ABSTRACT We introduce an extension to existing Distributed Hierarchical Graph Neuron (DHGN) netwo... more ABSTRACT We introduce an extension to existing Distributed Hierarchical Graph Neuron (DHGN) network for 2-dimensional binary pattern recognition. The new form of DHGN network, termed as receptive field DHGN network (RF-DHGN) is a hybrid of a receptive field layer for 2D feature extraction, and one or more DHGN subnets for feature recognition. All inputs to the network, in the form of synaptic weights are automatically determined through mode-based activation function within the RF neurons. The proposed scheme minimizes the need for large number of neurons as compared to the normal DHGN scheme. Furthermore, the results of preliminary recognition tests indicate high recognition accuracy, similar to existing DHGN approach for distributed pattern recognition.
2014 International Conference on Computer and Information Sciences (ICCOINS), 2014
ABSTRACT This article proposes the usage of Vector Symbolic Architectures for implementing Hierar... more ABSTRACT This article proposes the usage of Vector Symbolic Architectures for implementing Hierarchical Graph Neuron. The adoption of VSA representation maintains previously reported properties and performance characteristics of HGN and further makes it suitable for implementation in distributed wireless sensor networks of tiny devices.
... One Shot Associative Memory Method for Distorted Pattern Recognition Asad I. Khan and Anang H... more ... One Shot Associative Memory Method for Distorted Pattern Recognition Asad I. Khan and Anang Hudaya Muhamad Amin Clayton School of IT, Monash University, Clayton 3168 VIC, Australia {Asad.Khan, Anang.Hudaya}@infotech.monash.edu.au Abstract. ...
2010 Ninth IEEE International Symposium on Network Computing and Applications, 2010
Existing data management schemes in clouds are mainly based on Google File System (GFS) and MapRe... more Existing data management schemes in clouds are mainly based on Google File System (GFS) and MapReduce. Problems arise when data partitioning among numerous available nodes therein. This research paper explores new methods of partitioning and distributing data, that is, resource virtualization in cloud computing. Loosely-coupled associative computing techniques, which have so far not been considered for clouds, can provide the
2011 11th International Conference on Intelligent Systems Design and Applications, 2011
Abstract Network intrusion detection has remained a field of rigorous research over the past few ... more Abstract Network intrusion detection has remained a field of rigorous research over the past few years. Advances in computing performance, in terms of processing power and storage, have allowed the use of resource-intensive intelligent algorithms, to detect intrusive ...
2011 11th International Conference on Hybrid Intelligent Systems (HIS), 2011
Multiple-feature implementation enables a holistic approach towards pattern recognition procedure... more Multiple-feature implementation enables a holistic approach towards pattern recognition procedure that takes into consideration all significant features, which represent a particular set of complex patterns, such as images and sensor readings. This intends to reduce the bias effect of selecting only a single feature for classification/recognition purposes. In this paper we demonstrate the effectiveness of this approach in comparison with
2012 12th International Conference on Control Automation Robotics & Vision (ICARCV), 2012
ABSTRACT In Wireless Sensor Networks (WSNs), using physically sensed data for accurate automated ... more ABSTRACT In Wireless Sensor Networks (WSNs), using physically sensed data for accurate automated decision making is challenging. In response to these challenges, a combined Genetic Algorithm (GA) and pattern recognition scheme (PR) is presented in this paper. The aim of the scheme is to reduce the exponential relationship between problem size and time complexity of GA for guiding robots using WSN. The PR scheme presented in this paper is called Cellular Weighted Pattern Recogniser (CWPR) that simplifies computations and communications for energy conservation and speeds up recognition by leveraging the parallel distributed processing capabilities of WSN. Additionally, CWPR solves the problem of dilation, translation, and rotation to provide efficient pattern recognition in energy constrained WSN environments. Combining CWPR with GA allows GA to learn from experience and solve similar problems in fewer number of generations. The experimental results show that the approach efficiently supports a variety of PR applications for WSN guided robots.
2008 Ninth International Conference on Parallel and Distributed Computing, Applications and Technologies, 2008
... Wireless Sensor Networks Anang Hudaya Muhamad Amin and Asad I. Khan Clayton School of Informa... more ... Wireless Sensor Networks Anang Hudaya Muhamad Amin and Asad I. Khan Clayton School of Information Technology Monash University Clayton, 3168 VIC Australia {Anang.Hudaya, Asad.Khan}@infotech.monash.edu.au Abstract ...
2012 International Conference on Computer & Information Science (ICCIS), 2012
ABSTRACT Reduced-Distributed Hierarchical Graph Neuron (R-DHGN) is a one-shot learning distribute... more ABSTRACT Reduced-Distributed Hierarchical Graph Neuron (R-DHGN) is a one-shot learning distributed associative memory algorithm for data classification, which reduces the computational complexity of existing recognition algorithms by distributing the recognition process into smaller processing clusters. This paper investigates an effect of unsupervised one-shot learning mechanism for data classification within a computational network. This computational network may represent a network of objects than can be deployed in the existing Internet-of-Things (IoT) environment that offers seamless connectivity between smart devices such as sensors. Our approach extends the pattern recognition capability of Distributed Hierarchical Graph Neuron (DHGN). The interprocess communications of DHGN scheme is significantly reduced, and preliminary results obtained from the series of comparative analyses with other established classifiers have indicated the capability of R-DHGN to produce one-shot classification technique using a lightweight recognition mechanism. Simple dataset of iris plants have been used to demonstrate such capability of R-DHGN.
Page 401. Single-Cycle Image Recognition Using an Adaptive Granularity Associative Memory Network... more Page 401. Single-Cycle Image Recognition Using an Adaptive Granularity Associative Memory Network Anang Hudaya Muhamad Amin and Asad I. Khan Clayton School of Information Technology, Monash University, Clayton, 3168 VIC, Australia {Anang. Hudaya, Asad. ...
ABSTRACT We introduce an extension to existing Distributed Hierarchical Graph Neuron (DHGN) netwo... more ABSTRACT We introduce an extension to existing Distributed Hierarchical Graph Neuron (DHGN) network for 2-dimensional binary pattern recognition. The new form of DHGN network, termed as receptive field DHGN network (RF-DHGN) is a hybrid of a receptive field layer for 2D feature extraction, and one or more DHGN subnets for feature recognition. All inputs to the network, in the form of synaptic weights are automatically determined through mode-based activation function within the RF neurons. The proposed scheme minimizes the need for large number of neurons as compared to the normal DHGN scheme. Furthermore, the results of preliminary recognition tests indicate high recognition accuracy, similar to existing DHGN approach for distributed pattern recognition.
2014 International Conference on Computer and Information Sciences (ICCOINS), 2014
ABSTRACT This article proposes the usage of Vector Symbolic Architectures for implementing Hierar... more ABSTRACT This article proposes the usage of Vector Symbolic Architectures for implementing Hierarchical Graph Neuron. The adoption of VSA representation maintains previously reported properties and performance characteristics of HGN and further makes it suitable for implementation in distributed wireless sensor networks of tiny devices.
... One Shot Associative Memory Method for Distorted Pattern Recognition Asad I. Khan and Anang H... more ... One Shot Associative Memory Method for Distorted Pattern Recognition Asad I. Khan and Anang Hudaya Muhamad Amin Clayton School of IT, Monash University, Clayton 3168 VIC, Australia {Asad.Khan, Anang.Hudaya}@infotech.monash.edu.au Abstract. ...
Uploads