In this study, we investigate several facets of a due-date quotation problem and develop a model considering jointly due-date quotation, order acceptance and base-stock decisions in a hybrid make-to-stock (MTS) / make-to-order (MTO) and multi-class system with lead time sensitive Poisson demand and exponentially distributed service times. We seek to maximize profit considering lateness penalties and holding costs in the model. We consider three alternative due-date quotation policies each having different properties in terms of due-date flexibility as well as the utilization of state information. In order to evaluate the value of due-date flexibility as well as state information, the performances of the optimal policy and alternative policies are evaluated for various performance measures under different operating conditions. We also discuss the benefit of joint pooling of inventory and capacity under optimal policy and an accept-all policy.

No description available.

In this thesis we investigate visual place categorization by combining successful global image descriptors with a method of visual attention in order to automatically detect meaningful objects for places. The idea behind this is to incorporate information about typical objects for place categorization without the need for tedious labelling of important objects. Instead, the applied attention mechanism is intended to find the objects a human observer would focus first, so that the algorithm can use their discriminative power to conclude the place category. Besides this object-based place categorization approach we employ the Gist and the Centrist descriptor as holistic image descriptors. To access the power of all these descriptors we employ SVM-DAS (discriminative accumulation scheme) for cue integration and furthermore smooth the output trajectory with a delayed Hidden Markov Model. For the classification of the variety of descriptors we present and evaluate several classification methods. Among them is a joint probability modelling approach with two approximations as well as a modified KNN classifier, AdaBoost and SVM. The latter two classifiers are enhanced for multi-class use with a probabilistic computation scheme which treats the individual classifiers as peers and not as a hierarchical sequence. We check and tweak the different descriptors and classifiers in extensive tests mainly with a dataset of six homes. After these experiments we extend the basic algorithm with further filtering and tracking methods and evaluate their influence on the performance. Finally, we also test our algorithm within a university environment and on a real robot within a home environment.

The work deals with an introduction to classification algorithms. It then divides classifiers into unary, binary and multi-class and describes the different types of classifiers. The work compares individual classifiers and their areas of use. For unary classifiers, practical examples and a list of used architectures are given in the work. The work contains a chapter focused on the comparison of the effects of hyper parameters on the quality of unary classification for individual architectures. Part of the submission is a practical example of reimplementation of the unary classifier.

Text classification is used in information extraction and retrieval from a given text, and text classification has been considered as an important step to manage a vast number of records given in digital form that is far-reaching and expanding. This thesis addresses patent document classification problem into fifteen different categories or classes, where some classes overlap with other classes for practical reasons. For the development of the classification model using machine learning techniques, useful features have been extracted from the given documents. The features are used to classify patent document as well as to generate useful tag-words. The overall objective of this work is to systematize NASA’s patent management, by developing a set of automated tools that can assist NASA to manage and market its portfolio of intellectual properties (IP), and to enable easier discovery of relevant IP by users. We have identified an array of methods that can be applied such as k-Nearest Neighbors (kNN), two variations of the Support Vector Machine (SVM) algorithms, and two tree based classification algorithms: Random Forest and J48. The major research steps in this work consist of filtering techniques for variable selection, information gain and feature correlation analysis, and training and testing potential models using effective classifiers. Further, the obstacles associated with the imbalanced data were mitigated by adding synthetic data wherever appropriate, which resulted in a superior SVM classifier based model.

One option for enabling mobility between 5G nodes is to use a set of area-fixed reference beams in the downlink direction from each node. To save power these reference beams should be turned on only on demand, i.e. only if a mobile needs it. An User Equipment (UE) moving out of a beam's coverage will require a switch from one beam to another, preferably without having to turn on all possible beams to find out which one is the best. This thesis investigates how to transform the beam selection problem into a format suitable for machine learning and how good such solutions are compared to baseline models. The baseline models considered were beam overlap and average Reference Signal Received Power (RSRP), both building beam-to-beam maps. Emphasis in the thesis was on handovers between nodes and finding the beam with the highest RSRP. Beam-hit-rate and RSRP-difference (selected minus best) were key performance indicators and were compared for different numbers of activated beams. The problem was modeled as a Multiple Output Regression (MOR) problem and as a Multi-Class Classification (MCC) problem. Both problems are possible to solve with the random forest model, which was the learning model of choice during this work. An Ericsson simulator was used to simulate and collect data from a seven-site scenario with 40 UEs. Primary features available were the current serving beam index and its RSRP. Additional features, like position and distance, were suggested, though many ended up being limited either by the simulated scenario or by the cost of acquiring the feature in a real-world scenario. Using primary features only, learned models' performance were equal to or worse than the baseline models' performance. Adding distance improved the performance considerably, beating the baseline models, but still leaving room for more improvements.

Récemment les techniques d'apprentissage automatique ont montré leurs capacité à identifier des catégories d'images à partir de descripteurs extrait de caractéristiques visuels des images. Face à la croissance du nombre d'images et du nombre de catégories à traiter, plusieurs techniques ont été proposées pour réduire à la fois le coût calculatoire des méthodes et l'investissement humain en terme de supervision. Dans cette thèse nous proposons deux méthodes qui ont pour objectif de traiter un grand nombre d'images et de catégories. Nous proposons tout d'abord une solution reposant sur le concepts de recherche interactive. Le protocole de recherche interactive propose d'établir un « dialogue » entre le système d'apprentissage et l'utilisateur afin de minimiser l'effort d'annotation. Nous avons voulu dans ces travaux proposer une solution de recherche interactive adaptée aux méthodes de boosting . Ces méthodes combinent des classifieurs faibles pour produire un classifieur plus fort. Nous avons proposé une méthode de boosting interactif pour la recherche dans les images qui fit l'objet de deux articles (RFIA 2010, ICPR 2010). Ces méthodes proposent notamment une nouvelle manière de construire l'ensemble des classifieurs faibles sélectionnables par le boosting. Dans un second temps nous nous sommes consacré plus particulièrement aux méthodes à noyaux dans un contexte d'apprentissage plus classique. Ces méthodes ont montré de très bon résultats mais le choix de la fonction noyau et son réglage reste un enjeux important. Dans ces travaux, nous avons mis en place une nouvelle méthode d'apprentissage de fonction noyau multi-classes pour la classification de grande base d'images. Nous avons choisie d'utiliser un frameworks inspiré des méthodes de boosting pour créer un noyau fort à partir d'une combinaison de noyau plus faible. Nous utilisons la dualité entre fonction noyau et espace induit pour construit un nouvelle espace de représentation des données plus adapté à la catégorisation. L'idée de notre méthode est de construire de manière optimale ce nouvel espace de représentation afin qu'il permette l'apprentissage d'un nouveau classifieur plus rapide et de meilleures qualités. Chaque donnée multimédia sera alors représentée dans cette espace sémantique en lieu et place de sa représentation visuelle. Pour reproduire une approche similaire à une méthode de boosting, nous utilisons une construction incrémentale où des noyaux faibles sont entraînés dans une direction déterminée par les erreurs de l'itération précédente. Ces noyaux sont combinés à un facteur de pondération près, calculé grâce à la résolution analytique d'un problème d'optimisation. Ces travaux se basent sur des fondements mathématiques et font l'objet d'expériences montrant son intérêt pratique par comparaison avec les méthodes les plus récentes de la littérature. Ceux-ci sont présentés dans deux articles à Esann 2012 et ICIP 2012 ainsi que dans une soumission à MTAP. / Recent machine learning techniques have demonstrated their capability for identifying image categories using image features. Among these techniques, Support Vector Machines (SVM)present the best results, particularly when they are associated with a kernel function. However, nowadays image categorization task is very challenging owing to the sizes of benchmark datasets and the number of categories to be classified. In such a context, lot of effort has to be put in the design of the kernel functions and underlying high-level features. In this talk, we propose a new method to learn a kernel function for image categorization in large image databases. Our learning method is made of two steps :first, a kernel is built and semantic features are deduced ; then each class is learn thanks to a standard SVM. We adopt a Boosting framework to design and combine weak kernel functions targeting an ideal kernel. We propose a new iterative algorithm inspired from Boosting, to create a strong kernel. The weak kernels are learn thanks to the duality between the kernel space and the semantic feature space. We show that our method actually builds mapping functions which turn the initial input space to a new feature space where categories are better classified. Furthermore, our algorithm benefits from Boosting process to learn this kernel with a complexity linear with the size of the training set. Experiments are carried out on popular benchmarks and databases to show the properties and behavior of the proposed method. On the PASCAL VOC2006 database, we compare our method to simple early fusion, and on the Oxford Flowers databases we show that our method outperforms the best MKL techniques of the literature.

AdaBoost.M1 has been successfully applied to improve the accuracy of a learning algorithm for multi-class classification problems. However, it assumes that the performance of each base classifier must be better than 1/2, and this may be hard to achieve in practice for a multi-class problem. A new algorithm called AdaBoost.MK only requiring base classifiers better than a random guessing (1/k) is thus designed. Early SVM-based multi-class classification algorithms work by splitting the original problem into a set of two-class sub-problems. The time and space required by these algorithms are very demanding. In order to have low time and space complexities, we develop a base classifier that integrates one-class SVMs with discriminant functions. In this study, a hybrid method that integrates AdaBoost.MK and one-class SVMs with improved discriminant functions as the base classifiers is proposed to solve a multi-class classification problem. Experimental results on data sets from UCI and Statlog show that the proposed approach outperforms many popular multi-class algorithms including support vector clustering and AdaBoost.M1 with one-class SVMs as the base classifiers.

In the thesis, we propose a dual-mode message delivery system with mechanisms of time constrained paging and multi-class message. The pairing decision depends on the effective pairing time defined by the system when a bluetooth device comes into service range. Within the constrained pairing time, central server can deliver a message to the bluetooth device directly without re-pairing. Otherwise, the bluetooth device has to be paired with an intermediate node before it can receive a message. In addition, we store the number of times that bluetooth devices can move into the service range into a data base in order to send multi-class messages to these bluetooth devices. To demonstrate the proposed schemes, we implement a central server on Linux system and intermediate nodes on Window Mobile platform. We also design control packets associated with their message formats. Control messages can be exchanged between the central server and the intermediate node by the control packets, and data messages can be transmitted in a heterogeneous network, consisting of bluetooth and Wi-Fi. Finally, we measured the time saved without using pairing procedure and also verified that the system can dynamically adjust the classes of messages according to the number of times that bluetooth devices enter to the service range.

In any health-care system, early identification of individuals who are most at risk of developing an illness is vital, not only to ensure that a patient is provided with the appropriate treatment, but also to avoid the considerable costs associated with unnecessary hospitalization. To achieve this goal there is a need for a breakthrough prediction method that is capable of dealing with a real world medical data which is inherently complex. In this study, we show how standard classification algorithms can be employed collectively to predict the length of stay in a hospital of a patient in the upcoming year, based on their medical history. Multiple classifiers are used to perform the prediction task, since real world medical data is significantly complex making the classification task very challenging. The data is voluminous, consists of wide range of class values some of which with a few instances, and it is highly unbalanced making the classification of minority classes very difficult. We propose two Sequential Ensemble Classification (SEC) schemes, one based on an ensemble of homogeneous classifiers, and a second based on a heterogeneous ensemble of classifiers, in three hierarchical granularity levels. The goal of using this system is to provide increased performance over the standard classifiers. This method is highly beneficial when dealing with complex data which is multi-class and highly unbalanced.