Measures in Sectorization Problems

Sectorization means dividing a whole into parts (sectors), a procedure that occurs in many contexts and applications, usually to achieve some goal or to facilitate an activity. The objective may be a better organization or simplification of a large problem into smaller sub-problems. Examples of applications are political districting and sales territory division. When designing/comparing sectors some characteristics such as contiguity, equilibrium and compactness are usually considered. This paper presents and describes new generic measures and proposes a new measure, desirability, connected with the idea of preference.

HIERARCHICAL DECOMPOSITION OF MULTICLASS PROBLEMS

Several popular Machine Learning techniques are originally designed for the solution of two-class problems. However, several classification problems have more than two classes. One approach to deal with multiclass problems using binary classifiers is to decompose the multiclass problem into multiple binary sub-problems disposed in a binary tree. This approach requires a binary partition of the classes for each node of the tree, which defines the tree structure. This paper presents two algorithms to determine the tree structure taking into account information collected from the used dataset. This approach allows the tree structure to be determined automatically for any multiclass dataset.

Multiple-Retrieval Case-Based Reasoning for Course Timetabling Problems

The structured representation of cases by attribute graphs in a Case-Based Reasoning (CBR) system for course timetabling has been the subject of previous research by the authors. In that system, the case base is organised as a decision tree and the retrieval process chooses those cases which are sub attribute graph isomorphic to the new case. The drawback of that approach is that it is not suitable for solving large problems. This paper presents a multiple-retrieval approach that partitions a large problem into small solvable sub-problems by recursively inputting the unsolved part of the graph into the decision tree for retrieval. The adaptation combines the retrieved partial solutions of all the partitioned sub-problems and employs a graph heuristic method to construct the whole solution for the new case. We present a methodology which is not dependant upon problem specific information and which, as such, represents an approach which underpins the goal of building more general timetabling systems. We also explore the question of whether this multiple-retrieval CBR could be an effective initialisation method for local search methods such as Hill Climbing, Tabu Search and Simulated Annealing. Significant results are obtained from a wide range of experiments. An evaluation of the CBR system is presented and the impact of the approach on timetabling research is discussed. We see that the approach does indeed represent an effective initialisation method for these approaches.

Multiple-retrieval case-based reasoning for course timetabling problems

The structured representation of cases by attribute graphs in a Case-Based Reasoning (CBR) system for course timetabling has been the subject of previous research by the authors. In that system, the case base is organised as a decision tree and the retrieval process chooses those cases which are sub attribute graph isomorphic to the new case. The drawback of that approach is that it is not suitable for solving large problems. This paper presents a multiple-retrieval approach that partitions a large problem into small solvable sub-problems by recursively inputting the unsolved part of the graph into the decision tree for retrieval. The adaptation combines the retrieved partial solutions of all the partitioned sub-problems and employs a graph heuristic method to construct the whole solution for the new case. We present a methodology which is not dependant upon problem specific information and which, as such, represents an approach which underpins the goal of building more general timetabling systems. We also explore the question of whether this multiple-retrieval CBR could be an effective initialisation method for local search methods such as Hill Climbing, Tabu Search and Simulated Annealing. Significant results are obtained from a wide range of experiments. An evaluation of the CBR system is presented and the impact of the approach on timetabling research is discussed. We see that the approach does indeed represent an effective initialisation method for these approaches.

Decision Support Methods for Global Optomization

Immobile location-allocation (LA) problems is a type of LA problem that consists in determining the service each facility should offer in order to optimize some criterion (like the global demand), given the positions of the facilities and the customers. Due to the complexity of the problem, i.e. it is a combinatorial problem (where is the number of possible services and the number of facilities) with a non-convex search space with several sub-optimums, traditional methods cannot be applied directly to optimize this problem. Thus we proposed the use of clustering analysis to convert the initial problem into several smaller sub-problems. By this way, we presented and analyzed the suitability of some clustering methods to partition the commented LA problem. Then we explored the use of some metaheuristic techniques such as genetic algorithms, simulated annealing or cuckoo search in order to solve the sub-problems after the clustering analysis

Learning wind fields with multiple kernels

This paper presents multiple kernel learning (MKL) regression as an exploratory spatial data analysis and modelling tool. The MKL approach is introduced as an extension of support vector regression, where MKL uses dedicated kernels to divide a given task into sub-problems and to treat them separately in an effective way. It provides better interpretability to non-linear robust kernel regression at the cost of a more complex numerical optimization. In particular, we investigate the use of MKL as a tool that allows us to avoid using ad-hoc topographic indices as covariables in statistical models in complex terrains. Instead, MKL learns these relationships from the data in a non-parametric fashion. A study on data simulated from real terrain features confirms the ability of MKL to enhance the interpretability of data-driven models and to aid feature selection without degrading predictive performances. Here we examine the stability of the MKL algorithm with respect to the number of training data samples and to the presence of noise. The results of a real case study are also presented, where MKL is able to exploit a large set of terrain features computed at multiple spatial scales, when predicting mean wind speed in an Alpine region.

Intelligent Internet Data Searching

In this thesis author approaches the problem of automated text classification, which is one of basic tasks for building Intelligent Internet Search Agent. The work discusses various approaches to solving sub-problems of automated text classification, such as feature extraction and machine learning on text sources. Author also describes her own multiword approach to feature extraction and pres-ents the results of testing this approach using linear discriminant analysis based classifier, and classifier combining unsupervised learning for etalon extraction with supervised learning using common backpropagation algorithm for multilevel perceptron.

Consumers’ brand experiences of magazine brands

Tässä kvalitatiivisessa pro gradu -työssä tutkittiin kuluttajien brändikokemuksia aikakauslehtien brändeistä. Päätutkimusongelmana oli kuinka kuluttaja kokee aikakauslehden brändin. Alaongelmia olivat seuraavat: mitä ovat brändikokemukset, mitkä ovat brändikokemuksen ulottuvuudet, mitkä konseptit liittyvät brändikokemukseen, kuka on vastuussa brändikokemusten luomisesta ja miten brändikokemuksia luodaan. Tässä tutkimuksessa on analysoitu yhtä avointa kysymystä: kerro meille kokemuksiasi X lehden brändistä. Vastausten perusteella vastaajista muodostettiin kahdeksan ryhmää: Uloskasvaneet, Internet-sivujen käyttäjät, Toistoon kyllästyneet, Vertaistukijat, Nostalgiset, Uskolliset, Tyytyväiset ja Luottavaiset.

Monibrändisen arkkitehtuurin johtaminen

Työ käsittelee brändiportfolion hallintaa monibrändisessä arkkitehtuurissa. Työn painopiste on erityisesti erillisbrändi -strategioissa.Tutkimuksen keskeisin ongelma voidaan muotoilla kysymykseen: Miten erityispiirteitä monibrändisen arkkitehtuurin johtaminen sisältää? Alaongelmia kuuluvat seuraavat kysymykset: Mitä brändiportfolion rakentamisessa tulee ottaa huomioon? Mitä etuja monibrändisellä arkkitehtuurilla tavoitellaan? Mitä haasteita monibrändisen arkkitehtuurin johtaminen pitää sisällään? Miten monibrändisen strategian onnistumista arvioidaan ja mitataan?

Support Vector Machines: Training and Applications

The Support Vector Machine (SVM) is a new and very promising classification technique developed by Vapnik and his group at AT&T Bell Labs. This new learning algorithm can be seen as an alternative training technique for Polynomial, Radial Basis Function and Multi-Layer Perceptron classifiers. An interesting property of this approach is that it is an approximate implementation of the Structural Risk Minimization (SRM) induction principle. The derivation of Support Vector Machines, its relationship with SRM, and its geometrical insight, are discussed in this paper. Training a SVM is equivalent to solve a quadratic programming problem with linear and box constraints in a number of variables equal to the number of data points. When the number of data points exceeds few thousands the problem is very challenging, because the quadratic form is completely dense, so the memory needed to store the problem grows with the square of the number of data points. Therefore, training problems arising in some real applications with large data sets are impossible to load into memory, and cannot be solved using standard non-linear constrained optimization algorithms. We present a decomposition algorithm that can be used to train SVM's over large data sets. The main idea behind the decomposition is the iterative solution of sub-problems and the evaluation of, and also establish the stopping criteria for the algorithm. We present previous approaches, as well as results and important details of our implementation of the algorithm using a second-order variant of the Reduced Gradient Method as the solver of the sub-problems. As an application of SVM's, we present preliminary results we obtained applying SVM to the problem of detecting frontal human faces in real images.

Optimal Methodology for Synchronized Scheduling of Parallel Station Assembly with Air Transportation

We present an optimal methodology for synchronized scheduling of production assembly with air transportation to achieve accurate delivery with minimized cost in consumer electronics supply chain (CESC). This problem was motivated by a major PC manufacturer in consumer electronics industry, where it is required to schedule the delivery requirements to meet the customer needs in different parts of South East Asia. The overall problem is decomposed into two sub-problems which consist of an air transportation allocation problem and an assembly scheduling problem. The air transportation allocation problem is formulated as a Linear Programming Problem with earliness tardiness penalties for job orders. For the assembly scheduling problem, it is basically required to sequence the job orders on the assembly stations to minimize their waiting times before they are shipped by flights to their destinations. Hence the second sub-problem is modelled as a scheduling problem with earliness penalties. The earliness penalties are assumed to be independent of the job orders.

Building binary-tree-based multiclass classifiers using separability measures

Various popular machine learning techniques, like support vector machines, are originally conceived for the solution of two-class (binary) classification problems. However, a large number of real problems present more than two classes. A common approach to generalize binary learning techniques to solve problems with more than two classes, also known as multiclass classification problems, consists of hierarchically decomposing the multiclass problem into multiple binary sub-problems, whose outputs are combined to define the predicted class. This strategy results in a tree of binary classifiers, where each internal node corresponds to a binary classifier distinguishing two groups of classes and the leaf nodes correspond to the problem classes. This paper investigates how measures of the separability between classes can be employed in the construction of binary-tree-based multiclass classifiers, adapting the decompositions performed to each particular multiclass problem. (C) 2010 Elsevier B.V. All rights reserved.

An approach for solving the lot-sizing problem of a market-driven foundry

Foundries can be found all over Brazil and they are very important to its economy. In 2008, a mixed integer-programming model for small market-driven foundries was published, attempting to minimize delivery delays. We undertook a study of that model. Here, we present a new approach based on the decomposition of the problem into two sub-problems: production planning of alloys and production planning of items. Both sub-problems are solved using a Lagrangian heuristic based on transferences. An important aspect of the proposed heuristic is its ability to take into account a secondary practice objective solution: the furnace waste. Computational tests show that the approach proposed here is able to generate good quality solutions that outperform prior results. Journal of the Operational Research Society (2010) 61, 108-114. doi:10.1057/jors.2008.151

Interior point algorithm for linear programming used in transmission network synthesis

This article presents a well-known interior point method (IPM) used to solve problems of linear programming that appear as sub-problems in the solution of the long-term transmission network expansion planning problem. The linear programming problem appears when the transportation model is used, and when there is the intention to solve the planning problem using a constructive heuristic algorithm (CHA), ora branch-and-bound algorithm. This paper shows the application of the IPM in a CHA. A good performance of the IPM was obtained, and then it can be used as tool inside algorithm, used to solve the planning problem. Illustrative tests are shown, using electrical systems known in the specialized literature. (C) 2005 Elsevier B.V. All rights reserved.

An active/reactive predispatch model incorporating ramp rate constraints solved by dual decomposition/lagrangian relaxation

The Predispatch model (PD) calculates a short-term generation policy for power systems. In this work a PD model is proposed that improves two modeling aspects generally neglected in the literature: voltage/reactive power constraints and ramp rate constraints for generating units. Reactive power constraints turn the PD into a non-linear problem and the ramp rate constraints couple the problem dynamically in time domain. The solution of the PD is turned into a harder task when such constraints are introduced. The dual decomposition/ lagrangian relaxation technique is used in the solution approach for handing dynamic constraints. As a result the PD is decomposed into a series of independent Optimal Power Flow (FPO) sub problems, in which the reactive power is represented in detail. The solution of the independent FPO is coordinated by means of Lagrange multipliers, so that dynamic constraints are iteratively satisfied. Comparisons between dispatch policies calculated with and without the representation of ramp rate constraints are performed, using the IEEE 30 bus test system. The results point-out the importance of representing such constraints in the generation dispatch policy. © 2004 IEEE.