Adaptive Markov chain Monte Carlo and Bayesian filtering for state space models

This thesis is concerned with the state and parameter estimation in state space models. The estimation of states and parameters is an important task when mathematical modeling is applied to many different application areas such as the global positioning systems, target tracking, navigation, brain imaging, spread of infectious diseases, biological processes, telecommunications, audio signal processing, stochastic optimal control, machine learning, and physical systems. In Bayesian settings, the estimation of states or parameters amounts to computation of the posterior probability density function. Except for a very restricted number of models, it is impossible to compute this density function in a closed form. Hence, we need approximation methods. A state estimation problem involves estimating the states (latent variables) that are not directly observed in the output of the system. In this thesis, we use the Kalman filter, extended Kalman filter, Gauss–Hermite filters, and particle filters to estimate the states based on available measurements. Among these filters, particle filters are numerical methods for approximating the filtering distributions of non-linear non-Gaussian state space models via Monte Carlo. The performance of a particle filter heavily depends on the chosen importance distribution. For instance, inappropriate choice of the importance distribution can lead to the failure of convergence of the particle filter algorithm. In this thesis, we analyze the theoretical Lᵖ particle filter convergence with general importance distributions, where p ≥2 is an integer. A parameter estimation problem is considered with inferring the model parameters from measurements. For high-dimensional complex models, estimation of parameters can be done by Markov chain Monte Carlo (MCMC) methods. In its operation, the MCMC method requires the unnormalized posterior distribution of the parameters and a proposal distribution. In this thesis, we show how the posterior density function of the parameters of a state space model can be computed by filtering based methods, where the states are integrated out. This type of computation is then applied to estimate parameters of stochastic differential equations. Furthermore, we compute the partial derivatives of the log-posterior density function and use the hybrid Monte Carlo and scaled conjugate gradient methods to infer the parameters of stochastic differential equations. The computational efficiency of MCMC methods is highly depend on the chosen proposal distribution. A commonly used proposal distribution is Gaussian. In this kind of proposal, the covariance matrix must be well tuned. To tune it, adaptive MCMC methods can be used. In this thesis, we propose a new way of updating the covariance matrix using the variational Bayesian adaptive Kalman filter algorithm.

Amino acid composition of parturient plasma, the intervillous space of the placenta and the umbilical vein of term newborn infants

The objective of the present study was to determine the levels of amino acids in maternal plasma, placental intervillous space and fetal umbilical vein in order to identify the similarities and differences in amino acid levels in these compartments of 15 term newborns from normal pregnancies and deliveries. All amino acids, except tryptophan, were present in at least 186% higher concentrations in the intervillous space than in maternal venous blood, with the difference being statistically significant. This result contradicted the initial hypothesis of the study that the plasma amino acid levels in the placental intervillous space should be similar to those of maternal plasma. When the maternal venous compartment was compared with the umbilical vein, we observed values 103% higher on the fetal side which is compatible with currently accepted mechanisms of active amino acid transport. Amino acid levels of the placental intervillous space were similar to the values of the umbilical vein except for proline, glycine and aspartic acid, whose levels were significantly higher than fetal umbilical vein levels (average 107% higher). The elevated levels of the intervillous space are compatible with syncytiotrophoblast activity, which maintain high concentrations of free amino acids inside syncytiotrophoblast cells, permitting asymmetric efflux or active transport from the trophoblast cells to the blood in the intervillous space. The plasma amino acid levels in the umbilical vein of term newborns probably may be used as a standard of local normality for clinical studies of amino acid profiles.

Lean Innovation in Large Companies: a Case of Implementation in R&D

The main objective of the study is to evaluate the impact of Lean Innovation management philosophy on the creativity potential of the large multinational enterprise. A theory of Lean Innovation indicates that the modern company in any industry can successfully combine both waste-decreasing approach and innovative potential promotion through creativity cultivation or, at least, preservation. The theoretical part of the work covers the main factors, pros and cons of Lean thinking and Innovation management separately, along with generalized new product development overview. While the modern international market becomes more accessible for entrepreneural initiatives, small enterprises and start-ups, large international corporations are more subject to adopt the Lean Innovation approach in both operational and product development sectors due to extended resources and capabilities. Moreover, a multinational enterprise is a highly probable pioneer in Lean innovation implementation. The empirical part of the thesis refers to a case of large European enterprise, operating in many markets around the globe, that currently undergoes innovation management adjustments and implementations in product development while already have related themselves with operational process optimization process through Lean thinking. A goal of the work is to understand what kind of difficulties and consequences a large international firm faces when dealing with Lean Innovation to improve own performance, if they can be sealed for generalized approach.

The impact of industrial context and institutional factors on innovation and internationalization of Russian companies

Innovation and internationalization nowadays play significant role in formation of the competitive advantage for the companies and it is important to explore these processes deeply and use their potential. Such parameters, as industrial context and institutional factors, are essential parts of the business life and their influence may be crucial for the company’s performance. The Master’s Thesis explores impact of industrial context and institutional factors on innovation and internationalization of Russian companies. This research is quantitative and based on the analysis of the relevant scientific literature and results of the survey, conducted among Russian manufacturing companies. Findings show positive influence of industrial context factor “international threat” and institutional environment factor “lack of resources and the market imperfections” on internationalization of innovative Russian companies.

Plasma amino acids in pregnancy, placental intervillous space and preterm newborn infants

Plasma amino acid levels have never been studied in the placental intervillous space of preterm gestations. Our objective was to determine the possible relationship between plasma amino acids of maternal venous blood (M), of the placental intervillous space (PIVS) and of the umbilical vein (UV) of preterm newborn infants. Plasma amino acid levels were analyzed by ion-exchange chromatography in M from 14 parturients and in the PIVS and UV of their preterm newborn infants. Mean gestational age was 34 ± 2 weeks, weight = 1827 ± 510 g, and all newborns were considered adequate for gestational age. The mean Apgar score was 8 and 9 at the first and fifth minutes. Plasma amino acid values were significantly lower in M than in PIVS (166%), except for aminobutyric acid. On average, plasma amino acid levels were significantly higher in UV than in M (107%) and were closer to PIVS than to M values, except for cystine and aminobutyric acid (P < 0.05). Comparison of the mean plasma amino acid concentrations in the UV of preterm to those of term newborn infants previously studied by our group showed no significant difference, except for proline (P < 0.05), preterm > term. These data suggest that the mechanisms of active amino acid transport are centralized in the syncytiotrophoblast, with their passage to the fetus being an active bidirectional process with asymmetric efflux. PIVS could be a reserve amino acid space for the protection of the fetal compartment from inadequate maternal amino acid variations.

Innovation process development and business process integration - a multiple case study

The main objective of this study was to find out the bases for innovation model formulation in an existing organization based on cases. Innovation processes can be analyzed based on their needs and based on their emphasis on the business model development or R&D. The research was conducted in energy sector within one company by utilizing its projects as cases for the study. It is typical for the field of business that development is slow, although the case company has put emphasis on its innovation efforts. Analysis was done by identifying the cases’ needs and comparing them. The results were that because of the variances in the needs of the cases, the applicability of innovation process models varies. It was discovered that by dividing the process into two phases, a uniform model could be composed. This model would fulfill the needs of the cases and potential future projects as well.

Novel retrograde puncture method to establish preperitoneal space for laparoscopic direct inguinal hernia repair with internal ring suturing

The aim of this study was to explore the clinical efficacy of a novel retrograde puncture approach to establish a preperitoneal space for laparoscopic direct inguinal hernia repair with inguinal ring suturing. Forty-two patients who underwent laparoscopic inguinal hernia repair with retrograde puncture for preperitoneal space establishment as well as inguinal ring suturing between August 2013 and March 2014 at our hospital were enrolled. Preperitoneal space was successfully established in all patients, with a mean establishment time of 6 min. Laparoscopic repairs were successful in all patients, with a mean surgical time of 26±15.1 min. Mean postoperative hospitalization duration was 3.0±0.7 days. Two patients suffered from postoperative local hematomas, which were relieved after puncturing and drainage. Four patients had short-term local pain. There were no cases of chronic pain. Patients were followed up for 6 months to 1 year, and no recurrence was observed. Our results demonstrate that preperitoneal space established by the retrograde puncture technique can be successfully used in adult laparoscopic hernioplasty to avoid intraoperative mesh fixation, and thus reduce medical costs.

Innovation Process and Organizational Adoption of Disruptive Innovation: Empirical Evidence from Finnish Bio-Economy Companies.

The thesis aims to build a coherent view and understanding of the innovation process and organizational technology adoption in Finnish bio-economy companies with a focus on innovations of a disruptive nature. Disruptive innovations are exceptional hence in order to create generalizations and a unified view of the subject the perspective is also on less radical innovations. Other interests of the thesis are how ideas are discovered and generated and how the nature of the innovation and size of the company affect the technology adoption and innovation process. The data was collected by interviewing six small and six large Finnish bio-economy companies. The results suggest companies regardless of size consider innovation as a core asset in the competitive markets. Organizations want to be considered innovators and early adopters yet these qualities are limited by certain, mainly resource-based factors. In addition the industry, scalability and Finland’s geographical location when seeking funding provide certain challenges. The innovation process may be considered relatively similar whether the idea or technology stems from an internal or external source suggesting the technology adoption process can in fact be linked to the innovation process theories. Thus the thesis introduces a new theoretical model which based on the results of the study and the theories of technology adoption and innovation process aims on characterizing how ideas and technology from both external and internal sources generate into innovations. The innovation process is in large bio-economy companies most often similar to or a modified version of the stage-gate model, while small companies generally have less structured processes. Nevertheless the more disruptive the innovation, the less it fits in the structured processes. This implies disruptive innovation cannot be put in a certain mould but it is rather processed case-by-case.

Gearing Towards Omnichannel Grocery Retail Through Business Model Innovation

Technological innovations and the advent of digitalization have led retail business into one of its biggest transformations of all time. Consumer behaviour has changed rapidly and the customers are ever more powerful, demanding, tech-savvy and moving on various plat-forms. These attributes will continue to drive the development and robustly restructure the architecture of value creation in the retail business. The largest retail category, grocery yet awaits for a real disruption, but the signals for major change are already on the horizon. The first wave of online grocery retail was introduced in the mid 1990’s and it throve until millennium. Many overreactions, heavy investments and the burst IT-bubble almost stag-nated the whole industry for a long period of time. The second wave started with a venge-ance around 2010. Some research was carried out during the first wave from a single-viewpoint of online grocery retail, but without a comprehensive approach to online-offline business model integration. Now the accelerating growth of e-business has initiated an increased interest to examine the transformation from traditional business models towards e-business models and their integration on the companies’ traditional business models. This research strove to examine how can we recognize and analyze how digitalization and online channels are affecting the business models of grocery retail, by using business mod-el canvas as an analysis tool. Furthermore business model innovation and omnichannel retail were presented and suggested as potential solutions for these changes. 21 experts in online grocery industry were being interviewed. The thoughts of the informants were being qualitatively analysed by using an analysis tool called the business model canvas. The aim of this research was to portray a holistic view on the Omnichannel grocery retail business model, and the value chain, in which the case company Arina along with its partners are operating. The key conclusions exhibited that online grocery retail business model is not an alterna-tive model nor a substitute for the traditional grocery retail business model, though all of the business model elements are to some extent affected by it, but rather a complementary business model that should be integrated into the prevailing, conventional grocery retail business model. A set of business model elements, such as value proposition and distribu-tion channels were recognized as the most important ones and sources of innovation within these components were being illustrated. Segments for online grocery retail were empiri-cally established as polarized niche markets in contrast of the segmented mass-market of the conventional grocery retail. Business model innovation was proven to be a considera-ble method and a conceptual framework, by which to come across with new value proposi-tions that create competitive advantage for the company in the contemporary, changing business environment. Arina as a retailer can be considered as a industry model innovator, since it has initiated an entire industry in its market area, where other players have later on embarked on, and in which the contributors of the value chain, such as Posti depend on it to a great extent. Consumer behaviour clearly affects and appears everywhere in the digi-talized grocery trade and it drives customers to multiple platforms where retailers need to be present. Omnichannel retail business model was suggested to be the solution, in which the new technologies are being utilized, contemporary consumer behaviour is embedded in decision-making and all of the segments and their value propositions are being served seamlessly across the channels.

Challenges of user-driven innovations on late stages of innovation process: evidence from ICT companies

The goal of this thesis is to study user-driven innovations and user involvement throughout the innovation process in context of B2B companies. Significant emphasis in the analysis put onto the late stages of innovation process and commercialization of innovations. Thesis includes detailed review of theoretical concepts and underlying frameworks of innovation process, lead users and user-driven innovations. The empirical part of the thesis consist of interviews of the four companies from ICT industry, followed by the comprehensive analysis and comparison of the results. The presented findings indicate common challenges, which ICT companies face, when shifting towards innovation by users paradigm. Linkages and connections among current situation and theoretical frameworks presented in the discussion part of the thesis allow to draw practical managerial implications. The results of the research emphasize valuable insights and challenges of user interactions within innovation process as well as output and participation related benefits for the companies and users. The research points out current state of the user involvement techniques and tools used for user interactions as well as suggests the possibilities for improvement in the future.

Guidelines for small embedded system companies aiming to enter the space industry

This guide summarizes useful information about the European Space Agency (ESA), the European space industry, the ECSS standards and product assurance for small and medium enterprises that are aiming to enter the industry. Additionally, the applicability of agile development in space projects is discussed.