A feature selection approach for unsupervised classification based on clustering
- Authors: Rubinov, Alex , Soukhoroukova, Nadejda , Ugon, Julien
- Date: 2004
- Type: Text , Conference paper
- Relation: Paper presented at Sixth International Conference on Optimization: Techniques and Applications (ICOTA) , University of Ballarat, Ballarat, Victoria : 9th-11th December 2004
- Full Text: false
- Description: Data have been collected for many years in different scientific (industrial, medical) research groups. Very often these groups kept all the the they could collect. It is possible that the data contains a lot of noisy features which do not bring any information, but make the problem more complicated. The additional study of eliminating non-informative and selecting informative features is very important in the area of Data Mining. There are several feature selection methods which were developed for supervised classification. The area of feature selection for unsupervised classification is not so developed. In this paper we present a new feature selection approach for unsupervised classification, based on clustering and nonsmooth optimisation techniques.
- Description: 2003004085
A generalized subgradient method with piecewise linear subproblem
- Authors: Bagirov, Adil , Ganjehlou, Asef Nazari , Tor, Hakan , Ugon, Julien
- Date: 2010
- Type: Text , Journal article
- Relation: Dynamics of Continuous, Discrete and Impulsive Systems Series B: Applications and Algorithms Vol. 17, no. 5 (2010), p. 621-638
- Full Text: false
- Reviewed:
- Description: In this paper, a new version of the quasisecant method for nonsmooth nonconvex optimization is developed. Quasisecants are overestimates to the objective function in some neighborhood of a given point. Subgradients are used to obtain quasisecants. We describe classes of nonsmooth functions where quasisecants can be computed explicitly. We show that a descent direction with suffcient decrease must satisfy a set of linear inequalities. In the proposed algorithm this set of linear inequalities is solved by applying the subgradient algorithm to minimize a piecewise linear function. We compare results of numerical experiments between the proposed algorithm and subgradient method. Copyright © 2010 Watam Press.
Characterization theorem for best linear spline approximation with free knots
- Authors: Sukhorukova, Nadezda , Ugon, Julien
- Date: 2010
- Type: Text , Journal article
- Relation: Dynamics of Continuous, Discrete and Impulsive Systems Series B: Applications and Algorithms Vol. 17, no. 5 (2010), p. 687-708
- Full Text: false
- Reviewed:
- Description: A necessary condition for a best Chebyshev approximation by piecewise linear functions is derived using quasidifferential calculus. We first discover some properties of the knots joining the linear functions. Then we use these properties to obtain the optimality condition. This condition is stronger than existing results. We present an example of linear spline approximation where the existing optimality conditions are satisfied, but not the proposed one, which shows that it is not optimal. Copyright © 2010 Watam Press.
Workload coverage through nonsmooth optimization
- Authors: Sukhorukova, Nadezda , Ugon, Julien , Yearwood, John
- Date: 2009
- Type: Text , Journal article
- Relation: Optimization Methods and Software Vol. 24, no. 2 (2009), p. 285-298
- Full Text: false
- Reviewed:
- Description: In this paper, workload coverage is the problem of identifying a pattern of days worked and days off, along with the number of hours worked on each work day. This pattern must satisfy certain work-related constraints and fit best to a predefined workload. In our study, we formulate the problem of workload coverage as an optimization problem. We propose a number of models which take into consideration various staffing constraints. For each of these models, our study aims to find a compromise between an accurate workload coverage and the ability to solve the corresponding optimization problems in a reasonable time. Numerical experiments on each model are carried out and the results are presented. Interestingly, the nonlinear programming approaches are found to be competitive with linear programming ones. © 2009 Taylor & Francis.
An efficient algorithm for the incremental construction of a piecewise linear classifier
- Authors: Bagirov, Adil , Ugon, Julien , Webb, Dean
- Date: 2011
- Type: Text , Journal article
- Relation: Information Systems Vol. 36, no. 4 (2011), p. 782-790
- Relation: http://purl.org/au-research/grants/arc/DP0666061
- Full Text: false
- Reviewed:
- Description: In this paper the problem of finding piecewise linear boundaries between sets is considered and is applied for solving supervised data classification problems. An algorithm for the computation of piecewise linear boundaries, consisting of two main steps, is proposed. In the first step sets are approximated by hyperboxes to find so-called "indeterminate" regions between sets. In the second step sets are separated inside these "indeterminate" regions by piecewise linear functions. These functions are computed incrementally starting with a linear function. Results of numerical experiments are reported. These results demonstrate that the new algorithm requires a reasonable training time and it produces consistently good test set accuracy on most data sets comparing with mainstream classifiers. © 2010 Elsevier B.V. All rights reserved.
Solving Euclidian travelling salesman problem using discrete-gradient based clustering and kohonen neural network
- Authors: Ghosh, Moumita , Ugon, Julien , Ghosh, Ranadhir , Bagirov, Adil
- Date: 2004
- Type: Text , Conference paper
- Relation: Paper presented at ICOTA6: 6th International Conference on Optimization - Techniques and Applications, Ballarat, Victoria : 9th December, 2004
- Full Text: false
- Reviewed:
- Description: E1
- Description: 2003000864
Optimisation of operations of a water distribution system for reduced power usage
- Authors: Bagirov, Adil , Ugon, Julien , Barton, Andrew , Briggs, Steven
- Date: 2008
- Type: Text , Conference paper
- Relation: Paper presented at 9th National Conference on Hydraulics in Water Engineering: Hydraulics 2008, Darwin, Northern Territory : 22nd-26th September 2008
- Full Text: false
- Description: There are many improvements to operation that can be made to a water distribution system once it has been constructed and placed in ground. Pipes and associated storages and pumps are typically designed to meet average peak daily demands, offer some capacity for growth, and also allow for some deterioration of performance over time. However, the 'as constructed' performance of the pipeline is invariably different to what was designed on paper, and this is particularly so for anything other than design flows, such as during times of water restrictions when there are significantly reduced flows. Because of this, there remain significant benefits to owners and operators for the adaptive and global optimisation of such systems. The present paper uses the Ouyen subsystem of the Northern Mallee Pipeline, in Victoria, as a case study for the development of an optimisation model. This has been done with the intent of using this model to reduce costs and provide better service to customers on this system. The Ouyen subsystem consists of 1600 km of trunk and distribution pipeline servicing an area of 456,000 Ha. The system includes 2 fixed speed pumps diverting water from the Murray River at Liparoo into two 150 ML balancing storages at Ouyen, 4 variable speed pumps feeding water from the balancing storages into the pipeline system, 2 variable speed pressure booster pumps and 5 town balancing storages. When considering all these components of the system, power consumption becomes an important part of the overall operation. The present paper considers a global optimisation model to minimise power consumption while maintaining reasonable performance of the system. The main components of the model are described including the network structure and the costs functions associated with the system. The final model presents the cost functions associated with the pump scheduling, including the penalties descriptions associated with maintaining appropriate storages levels and pressure bounds within the water distribution network.
- Description: 2003006758
Optimization in wireless local area network
- Authors: Kouhbor, Shahnaz , Ugon, Julien , Kruger, Alexander , Rubinov, Alex , Branch, Philip
- Date: 2004
- Type: Text , Conference paper
- Relation: Paper presented at ICOTA6: 6th International Conference on Optimization - Techniques and Applications, Ballarat, Victoria : 9th December, 2004
- Full Text: false
- Reviewed:
- Description: 2003000886
Optimization solvers and problem formulations for solving data clustering problems
- Authors: Ugon, Julien
- Date: 2007
- Type: Text , Journal article
- Relation: Pacific Journal of Optimization Vol. 3, no. 2 (2007), p. 387-397
- Full Text: false
- Reviewed:
- Description: A popular apprach for solving complex optimization problems is through relaxation: some constraints are removed in order to have a convex problem approximating the original problem. On the other hand, direct approaches for solving such problems are becoming increasingly powerful. This paper examines two cases drawn from data analysis, in order to compare the two techniques.
- Description: C1
- Description: 2003004937
Fast modified global k-means algorithm for incremental cluster construction
- Authors: Bagirov, Adil , Ugon, Julien , Webb, Dean
- Date: 2011
- Type: Text , Journal article
- Relation: Pattern Recognition Vol. 44, no. 4 (2011), p. 866-876
- Relation: http://purl.org/au-research/grants/arc/DP0666061
- Full Text: false
- Reviewed:
- Description: The k-means algorithm and its variations are known to be fast clustering algorithms. However, they are sensitive to the choice of starting points and are inefficient for solving clustering problems in large datasets. Recently, incremental approaches have been developed to resolve difficulties with the choice of starting points. The global k-means and the modified global k-means algorithms are based on such an approach. They iteratively add one cluster center at a time. Numerical experiments show that these algorithms considerably improve the k-means algorithm. However, they require storing the whole affinity matrix or computing this matrix at each iteration. This makes both algorithms time consuming and memory demanding for clustering even moderately large datasets. In this paper, a new version of the modified global k-means algorithm is proposed. We introduce an auxiliary cluster function to generate a set of starting points lying in different parts of the dataset. We exploit information gathered in previous iterations of the incremental algorithm to eliminate the need of computing or storing the whole affinity matrix and thereby to reduce computational effort and memory usage. Results of numerical experiments on six standard datasets demonstrate that the new algorithm is more efficient than the global and the modified global k-means algorithms. © 2010 Elsevier Ltd. All rights reserved.
Queueing programming models in telecommunication network maintenance
- Authors: Ugon, Julien , Jia, Long , Ouveysi, Iradj
- Date: 2003
- Type: Text , Conference paper
- Relation: Paper presented at the Symposium on Industrial Optimisation and the 9th Australian Optimisation Day, Perth : 30th September, 2002
- Full Text: false
- Reviewed:
- Description: E1
- Description: 2003000350
Codifferential method for minimizing nonsmooth DC functions
- Authors: Bagirov, Adil , Ugon, Julien
- Date: 2011
- Type: Text , Journal article
- Relation: Journal of Global Optimization Vol. 50, no. 1 (2011), p. 3-22
- Relation: http://purl.org/au-research/grants/arc/DP0666061
- Full Text: false
- Reviewed:
- Description: In this paper, a new algorithm to locally minimize nonsmooth functions represented as a difference of two convex functions (DC functions) is proposed. The algorithm is based on the concept of codifferential. It is assumed that DC decomposition of the objective function is known a priori. We develop an algorithm to compute descent directions using a few elements from codifferential. The convergence of the minimization algorithm is studied and its comparison with different versions of the bundle methods using results of numerical experiments is given. © 2010 Springer Science+Business Media, LLC.
An algorithm for clusterwise linear regression based on smoothing techniques
- Authors: Bagirov, Adil , Ugon, Julien , Mirzayeva, Hijran
- Date: 2014
- Type: Text , Journal article
- Relation: Optimization Letters Vol. 9, no. 2 (2014), p. 375-390
- Full Text: false
- Reviewed:
- Description: We propose an algorithm based on an incremental approach and smoothing techniques to solve clusterwise linear regression (CLR) problems. This algorithm incrementally divides the whole data set into groups which can be easily approximated by one linear regression function. A special procedure is introduced to generate an initial solution for solving global optimization problems at each iteration of the incremental algorithm. Such an approach allows one to find global or approximate global solutions to the CLR problems. The algorithm is tested using several data sets for regression analysis and compared with the multistart and incremental Spath algorithms.
A novel piecewise linear classifier based on polyhedral conic and max-min separabilities
- Authors: Bagirov, Adil , Ugon, Julien , Webb, Dean , Ozturk, Gurkan , Kasimbeyli, Refail
- Date: 2011
- Type: Text , Journal article
- Relation: TOP Vol.21, no.1 (2011), p. 1-22
- Full Text: false
- Reviewed:
- Description: In this paper, an algorithm for finding piecewise linear boundaries between pattern classes is developed. This algorithm consists of two main stages. In the first stage, a polyhedral conic set is used to identify data points which lie inside their classes, and in the second stage we exclude those points to compute a piecewise linear boundary using the remaining data points. Piecewise linear boundaries are computed incrementally starting with one hyperplane. Such an approach allows one to significantly reduce the computational effort in many large data sets. Results of numerical experiments are reported. These results demonstrate that the new algorithm consistently produces a good test set accuracy on most data sets comparing with a number of other mainstream classifiers. © 2011 Sociedad de EstadÃstica e Investigación Operativa.
Compiling and using input-output frameworks through collaborative virtual laboratories
- Authors: Lenzen, Manfred , Geschke, Arne , Wiedmann, Thomas , Lane, Joe , Anderson, Neal , Baynes, Timothy , Boland, John , Daniels, Peter , Dey, Christopher , Fry, Jacob , Hadjikakou, Michalis , Kenway, Steven , Malik, Arunima , Moran, Daniel , Murray, Joy , Nettleton, Stuart , Poruschi, Lavinia , Reynolds, Christian , Rowley, Hazel , Ugon, Julien , Webb, Dean , West, James
- Date: 2014
- Type: Text , Journal article
- Relation: Science of the Total Environment Vol. 485-486, no. 1 (July 2014), p. 241-251
- Full Text: false
- Reviewed:
- Description: Compiling, deploying and utilising large-scale databases that integrate environmental and economic data have traditionally been labour- and cost-intensive processes, hindered by the large amount of disparate and misaligned data that must be collected and harmonised. The Australian Industrial Ecology Virtual Laboratory (IELab) is a novel, collaborative approach to compiling large-scale environmentally extended multi-region input-output (MRIO) models.The utility of the IELab product is greatly enhanced by avoiding the need to lock in an MRIO structure at the time the MRIO system is developed. The IELab advances the idea of the "mother-daughter" construction principle, whereby a regionally and sectorally very detailed "mother" table is set up, from which "daughter" tables are derived to suit specific research questions. By introducing a third tier - the "root classification" - IELab users are able to define their own mother-MRIO configuration, at no additional cost in terms of data handling. Customised mother-MRIOs can then be built, which maximise disaggregation in aspects that are useful to a family of research questions.The second innovation in the IELab system is to provide a highly automated collaborative research platform in a cloud-computing environment, greatly expediting workflows and making these computational benefits accessible to all users.Combining these two aspects realises many benefits. The collaborative nature of the IELab development project allows significant savings in resources. Timely deployment is possible by coupling automation procedures with the comprehensive input from multiple teams. User-defined MRIO tables, coupled with high performance computing, mean that MRIO analysis will be useful and accessible for a great many more research applications than would otherwise be possible. By ensuring that a common set of analytical tools such as for hybrid life-cycle assessment is adopted, the IELab will facilitate the harmonisation of fragmented, dispersed and misaligned raw data for the benefit of all interested parties.
Classification through incremental max-min separability
- Authors: Bagirov, Adil , Ugon, Julien , Webb, Dean , Karasozen, Bulent
- Date: 2011
- Type: Text , Journal article
- Relation: Pattern Analysis and Applications Vol. 14, no. 2 (2011), p. 165-174
- Relation: http://purl.org/au-research/grants/arc/DP0666061
- Full Text: false
- Reviewed:
- Description: Piecewise linear functions can be used to approximate non-linear decision boundaries between pattern classes. Piecewise linear boundaries are known to provide efficient real-time classifiers. However, they require a long training time. Finding piecewise linear boundaries between sets is a difficult optimization problem. Most approaches use heuristics to avoid solving this problem, which may lead to suboptimal piecewise linear boundaries. In this paper, we propose an algorithm for globally training hyperplanes using an incremental approach. Such an approach allows one to find a near global minimizer of the classification error function and to compute as few hyperplanes as needed for separating sets. We apply this algorithm for solving supervised data classification problems and report the results of numerical experiments on real-world data sets. These results demonstrate that the new algorithm requires a reasonable training time and its test set accuracy is consistently good on most data sets compared with mainstream classifiers. © 2010 Springer-Verlag London Limited.
A modified parallel optimization system for updating large-size time-evolving flow matrix
- Authors: Yu, Ting , Ugon, Julien , Yu, Wei
- Date: 2011
- Type: Text , Journal article
- Relation: Information Sciences Vol.194, no. (2011), p.57-67
- Full Text: false
- Reviewed:
- Description: Flow matrices are widely used in many disciplines, but few methods can estimate them. This paper presents a knowledge-based system as capable of estimating and updating large-size time-evolving flow matrix. The system in this paper consists of two major components with the purposes of matrix estimation and parallel optimization. The matrix estimation algorithm interprets and follows users' query scripts, retrieves data from various sources and integrates them for the matrix estimation. The parallel optimization component is built upon a supercomputing facility to utilize its computational power to efficiently process a large amount of data and estimate a large-size complex matrix. The experimental results demonstrate its outstanding performance and the acceptable accuracy by directly and indirectly comparing the estimation matrix with the actual matrix constructed by surveys. © 2011 Elsevier Inc. All rights reserved.
Patient admission prediction using a pruned fuzzy min-max neural network with rule extraction
- Authors: Wang, Jin , Lim, Cheepeng , Creighton, Douglas , Khorsavi, Abbas , Nahavandi, Saeid , Ugon, Julien , Vamplew, Peter , Stranieri, Andrew , Martin, Laura , Freischmidt, Anton
- Date: 2015
- Type: Text , Journal article
- Relation: Neural Computing and Applications Vol. 26, no. 2 (2015), p. 277-289
- Full Text: false
- Reviewed:
- Description: A useful patient admission prediction model that helps the emergency department of a hospital admit patients efficiently is of great importance. It not only improves the care quality provided by the emergency department but also reduces waiting time of patients. This paper proposes an automatic prediction method for patient admission based on a fuzzy min–max neural network (FMM) with rules extraction. The FMM neural network forms a set of hyperboxes by learning through data samples, and the learned knowledge is used for prediction. In addition to providing predictions, decision rules are extracted from the FMM hyperboxes to provide an explanation for each prediction. In order to simplify the structure of FMM and the decision rules, an optimization method that simultaneously maximizes prediction accuracy and minimizes the number of FMM hyperboxes is proposed. Specifically, a genetic algorithm is formulated to find the optimal configuration of the decision rules. The experimental results using a large data set consisting of 450740 real patient records reveal that the proposed method achieves comparable or even better prediction accuracy than state-of-the-art classifiers with the additional ability to extract a set of explanatory rules to justify its predictions.
Nonsmooth nonconvex optimization approach to clusterwise linear regression problems
- Authors: Bagirov, Adil , Ugon, Julien , Mirzayeva, Hijran
- Date: 2013
- Type: Text , Journal article
- Relation: European Journal of Operational Research Vol. 229, no. 1 (2013), p. 132-142
- Full Text: false
- Reviewed:
- Description: Clusterwise regression consists of finding a number of regression functions each approximating a subset of the data. In this paper, a new approach for solving the clusterwise linear regression problems is proposed based on a nonsmooth nonconvex formulation. We present an algorithm for minimizing this nonsmooth nonconvex function. This algorithm incrementally divides the whole data set into groups which can be easily approximated by one linear regression function. A special procedure is introduced to generate a good starting point for solving global optimization problems at each iteration of the incremental algorithm. Such an approach allows one to find global or near global solution to the problem when the data sets are sufficiently dense. The algorithm is compared with the multistart Späth algorithm on several publicly available data sets for regression analysis. © 2013 Elsevier B.V. All rights reserved.
- Description: 2003011018
Global optimality conditions and optimization methods for polynomial programming problems
- Authors: Wu, Zhiyou , Tian, Jing , Ugon, Julien
- Date: 2015
- Type: Text , Journal article
- Relation: Journal of Global Optimization Vol. 62, no. 4 (2015), p. 617-641
- Full Text: false
- Reviewed:
- Description: This paper is concerned with the general polynomial programming problem with box constraints, including global optimality conditions and optimization methods. First, a necessary global optimality condition for a general polynomial programming problem with box constraints is given. Then we design a local optimization method by using the necessary global optimality condition to obtain some strongly or -strongly local minimizers which substantially improve some KKT points. Finally, a global optimization method, by combining the new local optimization method and an auxiliary function, is designed. Numerical examples show that our methods are efficient and stable.