Purpose - Given the existing obstacles to succeed with the front end of innovation, and the possibility to learn from the operations of hidden champions with their relative high rate of success in the market, this study aims to investigate how Nordic small and medium-sized hidden champions manage their front end of innovation. Method - The study is a multiple case study that uses an exploratory approach. A total of 19 interviews were conducted in two waves and across eleven organizations from the Nordic region. The collected data was analyzed through a thematic analysis. Findings - The thematic analysis resulted in two themes: Innovation Management and Stakeholder Management, that each centralize on different aspects managed. Central to the findings was the involvement of customers and external organizations into the front end of the innovation process. The findings built the foundation to a framework on enabling factors for the front end of innovation and thereby how other organizations should improve their front end of innovation. Moreover, a process framework is also presented which conceptualizes how organizations should operate their front end of innovation efforts. Theoretical contribution - This study adds to the literature by providing insights on how hidden champions manage their front end of innovation, confirming findings from other regions and adding to existing literature on the importance of involving customers in the front end of innovation process. Moreover, the study conceptualizes the factors that enable hidden champions to succeed in their front end of innovation. Managerial contribution - This study provides an overview of how managers can structure their front end of innovation process to increase the chances of successful innovation management. Limitations and future research - This study focuses on small and medium-sized hidden champions, thus the generalizability of this study may not apply to larger organizations. Therefore future research is proposed for large hidden champions to examine if the findings in this study are applicable for them. This study also examines hidden champions in a B2B context, therefore future research should be conducted regarding hidden champions in B2C to examine the applicability of our findings in that field.

This thesis investigates the application of Hidden Markov Models (HMMs) to model financial volatility-regimes and presents a parameter learning approach using real-world data. Although HMMs as regime-switching models are established, empirical studies regarding the parameter estimation of such models remain limited. We address this issue by creating a systematic approach (algorithm) for parameter learning using Python programming and the hmmlearn library. The algorithm works by initializing a wide range of random parameter values for an HMM and maximizing the log-likelihood of an observation sequence, obtained from market data, using expectation-maximization; the optimal number of volatility regimes for the HMM is determined using information criterion. By training models on historical market and volatility index data, we found that a discrete model is favored for volatility modeling and option pricing due to its low complexity and high customizability, and a Gaussian model is favored for asset allocation and price simulation due to its ability to model market regimes. However, practical applications of these models were not researched, and thus, require further studies to test and calibrate.

Le premier chapitre présente les axes principaux de recherche ainsi que les problèmes traités dans cette thèse. Plus précisément, il expose une synthèse sur le sujet, en y donnant les propriétés essentielles pour la bonne compréhension de cette étude, accompagnée des références bibliographiques les plus importantes. Il présente également les motivations de ce travail en précisant les contributions originales dans ce domaine. Le deuxième chapitre est composé d’une recherche originale sur l’estimation du risque sismique, dans la zone du nord de la mer Egée (Grèce), en faisant usage de la théorie des processus semi-markoviens à temps continue. Il propose des estimateurs des mesures importantes qui caractérisent les processus semi-markoviens, et fournit une modélisation dela prévision de l’instant de réalisation d’un séisme fort ainsi que la probabilité et la grandeur qui lui sont associées. Les chapitres 3 et 4 comprennent une première tentative de modélisation du processus de génération des séismes au moyen de l’application d’un temps discret des modèles cachés markoviens et semi-markoviens, respectivement. Une méthode d’estimation non paramétrique est appliquée, qui permet de révéler des caractéristiques fondamentales du processus de génération des séismes, difficiles à détecter autrement. Des quantités importantes concernant les niveaux des tensions sont estimées au moyen des modèles proposés. Le chapitre 5 décrit les résultats originaux du présent travail à la théorie des processus stochastiques, c’est- à-dire l’étude et l’estimation du « Intensité du temps d’entrée en temps discret (DTIHT) » pour la première fois dans des chaînes semi-markoviennes et des chaînes de renouvellement markoviennes cachées. Une relation est proposée pour le calcul du DTIHT et un nouvel estimateur est présenté dans chacun de ces cas. De plus, les propriétés asymptotiques des estimateurs proposés sont obtenues, à savoir, la convergence et la normalité asymptotique. Le chapitre 6 procède ensuite à une étude de comparaison entre le modèle markovien caché et le modèle semi-markovien caché dans un milieu markovien et semi-markovien en vue de rechercher d’éventuelles différences dans leur comportement stochastique, déterminé à partir de la matrice de transition de la chaîne de Markov (modèle markovien caché) et de la matrice de transition de la chaîne de Markov immergée (modèle semi-markovien caché). Les résultats originaux concernent le cas général où les distributions sont considérées comme distributions des temps de séjour ainsi que le cas particulier des modèles qui sont applique´s dans les chapitres précédents où les temps de séjour sont estimés de manière non-paramétrique. L’importance de ces différences est spécifiée à l’aide du calcul de la valeur moyenne et de la variance du nombre de sauts de la chaîne de Markov (modèle markovien caché) ou de la chaîne de Markov immergée (modèle semi-markovien caché) pour arriver dans un état donné, pour la première fois. Enfin, le chapitre 7 donne des conclusions générales en soulignant les points les plus marquants et des perspectives pour développements futurs. / The first chapter describes the definition of the subject under study, the current state of science in this area and the objectives. In the second chapter, continuous-time semi-Markov models are studied and applied in order to contribute to seismic hazard assessment in Northern Aegean Sea (Greece). Expressions for different important indicators of the semi- Markov process are obtained, providing forecasting results about the time, the space and the magnitude of the ensuing strong earthquake. Chapters 3 and 4 describe a first attempt to model earthquake occurrence by means of discrete-time hidden Markov models (HMMs) and hidden semi-Markov models (HSMMs), respectively. A nonparametric estimation method is followed by means of which, insights into features of the earthquake process are provided which are hard to detect otherwise. Important indicators concerning the levels of the stress field are estimated by means of the suggested HMM and HSMM. Chapter 5 includes our main contribution to the theory of stochastic processes, the investigation and the estimation of the discrete-time intensity of the hitting time (DTIHT) for the first time referring to semi-Markov chains (SMCs) and hidden Markov renewal chains (HMRCs). A simple formula is presented for the evaluation of the DTIHT along with its statistical estimator for both SMCs and HMRCs. In addition, the asymptotic properties of the estimators are proved, including strong consistency and asymptotic normality. In chapter 6, a comparison between HMMs and HSMMs in a Markov and a semi-Markov framework is given in order to highlight possible differences in their stochastic behavior partially governed by their transition probability matrices. Basic results are presented in the general case where specific distributions are assumed for sojourn times as well as in the special case concerning the models applied in the previous chapters, where the sojourn time distributions are estimated non-parametrically. The impact of the differences is observed through the calculation of the mean value and the variance of the number of steps that the Markov chain (HMM case) and the EMC (HSMM case) need to make for visiting for the first time a particular state. Finally, Chapter 7 presents concluding remarks, perspectives and future work.

Many real-world systems experience deterioration with usage and age, which often leads to low product quality, high production cost, and low system availability. Most previous maintenance and reliability models in the literature do not incorporate condition monitoring information for decision making, which often results in poor failure prediction for partially observable deteriorating systems. For that reason, the development of fault prediction and control scheme using condition-based maintenance techniques has received considerable attention in recent years. This research presents a new framework for predicting failures of a partially observable deteriorating system using Bayesian control techniques. A time series model is fitted to a vector observation process representing partial information about the system state. Residuals are then calculated using the fitted model, which are indicative of system deterioration. The deterioration process is modeled as a 3-state continuous-time homogeneous Markov process. States 0 and 1 are not observable, representing healthy (good) and unhealthy (warning) system operational conditions, respectively. Only the failure state 2 is assumed to be observable. Preventive maintenance can be carried out at any sampling epoch, and corrective maintenance is carried out upon system failure. The form of the optimal control policy that maximizes the long-run expected average availability per unit time has been investigated. It has been proved that a control limit policy is optimal for decision making. The model parameters have been estimated using the Expectation Maximization (EM) algorithm. The optimal Bayesian fault prediction and control scheme, considering long-run average availability maximization along with a practical statistical constraint, has been proposed and compared with the age-based replacement policy. The optimal control limit and sampling interval are calculated in the semi-Markov decision process (SMDP) framework. Another Bayesian fault prediction and control scheme has been developed based on the average run length (ARL) criterion. Comparisons with traditional control charts are provided. Formulae for the mean residual life and the distribution function of system residual life have been derived in explicit forms as functions of a posterior probability statistic. The advantage of the Bayesian model over the well-known 2-parameter Weibull model in system residual life prediction is shown. The methodologies are illustrated using simulated data, real data obtained from the spectrometric analysis of oil samples collected from transmission units of heavy hauler trucks in the mining industry, and vibration data from a planetary gearbox machinery application.

Many real-world systems experience deterioration with usage and age, which often leads to low product quality, high production cost, and low system availability. Most previous maintenance and reliability models in the literature do not incorporate condition monitoring information for decision making, which often results in poor failure prediction for partially observable deteriorating systems. For that reason, the development of fault prediction and control scheme using condition-based maintenance techniques has received considerable attention in recent years. This research presents a new framework for predicting failures of a partially observable deteriorating system using Bayesian control techniques. A time series model is fitted to a vector observation process representing partial information about the system state. Residuals are then calculated using the fitted model, which are indicative of system deterioration. The deterioration process is modeled as a 3-state continuous-time homogeneous Markov process. States 0 and 1 are not observable, representing healthy (good) and unhealthy (warning) system operational conditions, respectively. Only the failure state 2 is assumed to be observable. Preventive maintenance can be carried out at any sampling epoch, and corrective maintenance is carried out upon system failure. The form of the optimal control policy that maximizes the long-run expected average availability per unit time has been investigated. It has been proved that a control limit policy is optimal for decision making. The model parameters have been estimated using the Expectation Maximization (EM) algorithm. The optimal Bayesian fault prediction and control scheme, considering long-run average availability maximization along with a practical statistical constraint, has been proposed and compared with the age-based replacement policy. The optimal control limit and sampling interval are calculated in the semi-Markov decision process (SMDP) framework. Another Bayesian fault prediction and control scheme has been developed based on the average run length (ARL) criterion. Comparisons with traditional control charts are provided. Formulae for the mean residual life and the distribution function of system residual life have been derived in explicit forms as functions of a posterior probability statistic. The advantage of the Bayesian model over the well-known 2-parameter Weibull model in system residual life prediction is shown. The methodologies are illustrated using simulated data, real data obtained from the spectrometric analysis of oil samples collected from transmission units of heavy hauler trucks in the mining industry, and vibration data from a planetary gearbox machinery application.

No description available.

Changes in gene expression programs play a central role in cancer. Chromosomal aberrations such as deletions, duplications and translocations of DNA segments can lead to highly significant positive correlations of gene expression levels of neighboring genes. This should be utilized to improve the analysis of tumor expression profiles. Here, we develop a novel model class of autoregressive higher-order Hidden Markov Models (HMMs) that carefully exploit local data-dependent chromosomal dependencies to improve the identification of differentially expressed genes in tumor. Autoregressive higher-order HMMs overcome generally existing limitations of standard first-order HMMs in the modeling of dependencies between genes in close chromosomal proximity by the simultaneous usage of higher-order state-transitions and autoregressive emissions as novel model features. We apply autoregressive higher-order HMMs to the analysis of breast cancer and glioma gene expression data and perform in-depth model evaluation studies. We find that autoregressive higher-order HMMs clearly improve the identification of overexpressed genes with underlying gene copy number duplications in breast cancer in comparison to mixture models, standard first- and higher-order HMMs, and other related methods. The performance benefit is attributed to the simultaneous usage of higher-order state-transitions in combination with autoregressive emissions. This benefit could not be reached by using each of these two features independently. We also find that autoregressive higher-order HMMs are better able to identify differentially expressed genes in tumors independent of the underlying gene copy number status in comparison to the majority of related methods. This is further supported by the identification of well-known and of previously unreported hotspots of differential expression in glioblastomas demonstrating the efficacy of autoregressive higher-order HMMs for the analysis of individual tumor expression profiles. Moreover, we reveal interesting novel details of systematic alterations of gene expression levels in known cancer signaling pathways distinguishing oligodendrogliomas, astrocytomas and glioblastomas.

Many complexity measures are defined as the size of a minimal representation in a specific model class. One such complexity measure, which is important because it is widely applied, is statistical complexity. It is defined for discrete-time, stationary stochastic processes within a theory called computational mechanics. Here, a mathematically rigorous, more general version of this theory is presented, and abstract properties of statistical complexity as a function on the space of processes are investigated. In particular, weak-* lower semi-continuity and concavity are shown, and it is argued that these properties should be shared by all sensible complexity measures. Furthermore, a formula for the ergodic decomposition is obtained. The same results are also proven for two other complexity measures that are defined by different model classes, namely process dimension and generative complexity. These two quantities, and also the information theoretic complexity measure called excess entropy, are related to statistical complexity, and this relation is discussed here. It is also shown that computational mechanics can be reformulated in terms of Frank Knight's prediction process, which is of both conceptual and technical interest. In particular, it allows for a unified treatment of different processes and facilitates topological considerations. Continuity of the Markov transition kernel of a discrete version of the prediction process is obtained as a new result.

This study has developed a tool for explaining why employees fail to speak up with regard to work related criticism; there is a hidden policy of silence that teaches employees to remain silent. This hidden policy is here designated as the "Curriculum Silentium" and is described in detail on the basis of empirical and theoretical data. After identifying a gap between the intentionally and experienced policy for employees freedom of speech in organizations I suggest that there are on-going unofficial, partially hidden learning processes in the organizations. The overall research question is; How does the Curriculum Silentium; the hidden policy of silence among employees, look like?  I make an analytic construction of the hidden policy as if it were planned policy, using the didactic categories applicable to organizations. These didactic categories are: goals, content, teaching strategies and the motivation of employees. The empirical data was collected in three different organizations: an elementary school, a home for the elderly and a factory in the process industry, using qualitative methods such as interviews and observation. The theoretical foundation of the study is taken from existing theory within the field of work life research and educational science. The study is not a comparative study of the three organizations, but does involve a comparison of whether and how the Curriculum Silentium is expressed in three such different organizations. The challenge of examining hidden relationships in organizations was met through the development of guidelines for an analytical approach called a critical didactic relations analysis. The study concludes that a hidden policy of silence resembling that presented here exists in organizations where employees fail to voice working life related criticism.

Techn. Univ., Diss.--München, 2002.