Bayesian methods for estimation, optimization and experimental design

Mathematical models often contain parameters that need to be calibrated from measured data. The emergence of efficient Markov Chain Monte Carlo (MCMC) methods has made the Bayesian approach a standard tool in quantifying the uncertainty in the parameters. With MCMC, the parameter estimation problem can be solved in a fully statistical manner, and the whole distribution of the parameters can be explored, instead of obtaining point estimates and using, e.g., Gaussian approximations. In this thesis, MCMC methods are applied to parameter estimation problems in chemical reaction engineering, population ecology, and climate modeling. Motivated by the climate model experiments, the methods are developed further to make them more suitable for problems where the model is computationally intensive. After the parameters are estimated, one can start to use the model for various tasks. Two such tasks are studied in this thesis: optimal design of experiments, where the task is to design the next measurements so that the parameter uncertainty is minimized, and model-based optimization, where a model-based quantity, such as the product yield in a chemical reaction model, is optimized. In this thesis, novel ways to perform these tasks are developed, based on the output of MCMC parameter estimation. A separate topic is dynamical state estimation, where the task is to estimate the dynamically changing model state, instead of static parameters. For example, in numerical weather prediction, an estimate of the state of the atmosphere must constantly be updated based on the recently obtained measurements. In this thesis, a novel hybrid state estimation method is developed, which combines elements from deterministic and random sampling methods.

Kernel-Based Ranking. Methods for Learning and Performance Estimation

Machine learning provides tools for automated construction of predictive models in data intensive areas of engineering and science. The family of regularized kernel methods have in the recent years become one of the mainstream approaches to machine learning, due to a number of advantages the methods share. The approach provides theoretically well-founded solutions to the problems of under- and overfitting, allows learning from structured data, and has been empirically demonstrated to yield high predictive performance on a wide range of application domains. Historically, the problems of classification and regression have gained the majority of attention in the field. In this thesis we focus on another type of learning problem, that of learning to rank. In learning to rank, the aim is from a set of past observations to learn a ranking function that can order new objects according to how well they match some underlying criterion of goodness. As an important special case of the setting, we can recover the bipartite ranking problem, corresponding to maximizing the area under the ROC curve (AUC) in binary classification. Ranking applications appear in a large variety of settings, examples encountered in this thesis include document retrieval in web search, recommender systems, information extraction and automated parsing of natural language. We consider the pairwise approach to learning to rank, where ranking models are learned by minimizing the expected probability of ranking any two randomly drawn test examples incorrectly. The development of computationally efficient kernel methods, based on this approach, has in the past proven to be challenging. Moreover, it is not clear what techniques for estimating the predictive performance of learned models are the most reliable in the ranking setting, and how the techniques can be implemented efficiently. The contributions of this thesis are as follows. First, we develop RankRLS, a computationally efficient kernel method for learning to rank, that is based on minimizing a regularized pairwise least-squares loss. In addition to training methods, we introduce a variety of algorithms for tasks such as model selection, multi-output learning, and cross-validation, based on computational shortcuts from matrix algebra. Second, we improve the fastest known training method for the linear version of the RankSVM algorithm, which is one of the most well established methods for learning to rank. Third, we study the combination of the empirical kernel map and reduced set approximation, which allows the large-scale training of kernel machines using linear solvers, and propose computationally efficient solutions to cross-validation when using the approach. Next, we explore the problem of reliable cross-validation when using AUC as a performance criterion, through an extensive simulation study. We demonstrate that the proposed leave-pair-out cross-validation approach leads to more reliable performance estimation than commonly used alternative approaches. Finally, we present a case study on applying machine learning to information extraction from biomedical literature, which combines several of the approaches considered in the thesis. The thesis is divided into two parts. Part I provides the background for the research work and summarizes the most central results, Part II consists of the five original research articles that are the main contribution of this thesis.

Integration of microgrids into electricity distribution networks

Existing electricity distribution system is under pressure because implementation of distributed generation changes the grid configuration and also because some customers demand for better distribution reliability. In a short term, traditional network planning does not offer techno-economical solutions for the challenges and therefore the idea of microgrids is introduced. Islanding capability of microgrids is expected to enable better reliability by reducing effects of faults. The aim of the thesis is to discuss challenges in integration of microgrids into distribution networks. Study discusses development of microgrid related smart grid features and gives estimation of the guideline of microgrid implementation. Thesis also scans microgrid pilots around the world and introduces the most relevant projects. Analysis reveals that the main focus of researched studies is on low voltage microgrids. This thesis extends the idea to medium voltage distribution system and introduces challenges related to medium voltage microgrid implementation. Differences of centralized and distributed microgrid models are analyzed and the centralized model is discovered to be easiest to implement into existing distribution system. Preplan of medium voltage microgrid pilot is also carried out in this thesis.

Testing Futures Market Efficiency and Optimal Hedge Ratio Estimation: Evidence from Futures and Options on RTS

This study investigates futures market efficiency and optimal hedge ratio estimation. First, cointegration between spot and futures prices is studied using Johansen method, with two different model specifications. If prices are found cointegrated, restrictions on cointegrating vector and adjustment coefficients are imposed, to account for unbiasedness, weak exogeneity and prediction hypothesis. Second, optimal hedge ratios are estimated using static OLS, and time-varying DVEC and CCC models. In-sample and out-of-sample results for one, two and five period ahead are reported. The futures used in thesis are RTS index, EUR/RUB exchange rate and Brent oil, traded in Futures and options on RTS.(FORTS) For in-sample period, data points were acquired from start of trading of each futures contract, RTS index from August 2005, EUR/RUB exchange rate March 2009 and Brent oil October 2008, lasting till end of May 2011. Out-of-sample period covers start of June 2011, till end of December 2011. Our results indicate that all three asset pairs, spot and futures, are cointegrated. We found RTS index futures to be unbiased predictor of spot price, mixed evidence for exchange rate, and for Brent oil futures unbiasedness was not supported. Weak exogeneity results for all pairs indicated spot price to lead in price discovery process. Prediction hypothesis, unbiasedness and weak exogeneity of futures, was rejected for all asset pairs. Variance reduction results varied between assets, in-sample in range of 40-85 percent and out-of sample in range of 40-96 percent. Differences between models were found small, except for Brent oil in which OLS clearly dominated. Out-of-sample results indicated exceptionally high variance reduction for RTS index, approximately 95 percent.

Optimization as a method of search engine marketing

This study is dedicated to search engine marketing (SEM). It aims for developing a business model of SEM firms and to provide explicit research of trustworthy practices of virtual marketing companies. Optimization is a general term that represents a variety of techniques and methods of the web pages promotion. The research addresses optimization as a business activity, and it explains its role for the online marketing. Additionally, it highlights issues of unethical techniques utilization by marketers which created relatively negative attitude to them on the Internet environment. Literature insight combines in the one place both technical and economical scientific findings in order to highlight technological and business attributes incorporated in SEM activities. Empirical data regarding search marketers was collected via e-mail questionnaires. 4 representatives of SEM companies were engaged in this study to accomplish the business model design. Additionally, the fifth respondent was a representative of the search engine portal, who provided insight on relations between search engines and marketers. Obtained information of the respondents was processed qualitatively. Movement of commercial organizations to the online market increases demand on promotional programs. SEM is the largest part of online marketing, and it is a prerogative of search engines portals. However, skilled users, or marketers, are able to implement long-term marketing programs by utilizing web page optimization techniques, key word consultancy or content optimization to increase web site visibility to search engines and, therefore, user’s attention to the customer pages. SEM firms are related to small knowledge-intensive businesses. On the basis of data analysis the business model was constructed. The SEM model includes generalized constructs, although they represent a wider amount of operational aspects. Constructing blocks of the model includes fundamental parts of SEM commercial activity: value creation, customer, infrastructure and financial segments. Also, approaches were provided on company’s differentiation and competitive advantages evaluation. It is assumed that search marketers should apply further attempts to differentiate own business out of the large number of similar service providing companies. Findings indicate that SEM companies are interested in the increasing their trustworthiness and the reputation building. Future of the search marketing is directly depending on search engines development.

Future trends of Finnish cement based construction

The construction material sector, as a capital intensive industry, is highly vulnerable to rapid fluctuations in the economic cycles. In Finland this was witnessed especially during the late 2000s, as in 2007 and 2008 the demand for several construction materials exceeded their supply and right after this, in 2009 the demand collapsed fast as a result of an international recession. These factors brought about the need to study the future trends of the market place of the commissioning company, Finnsementti Oy. As reliable short term market forecasts for the sector are difficult to compose, the study concentrates primarily in examining and identifying the trends that are likely to affect the Finnish cement industry, and as an extension, the concrete industry in a frame of 10 to 15 years. The study’s scope comprehends also the examination of the domestic construction sector, as it represents the end user industry of both cement and concrete. These motives for the study produce the research problem, which is to conduct a trend analysis for cement based building in the Finnish market area in the 2020s. The theoretical frame for composing a trend analysis in the case of this study is twofold. This is due to the fact that both, the macro and micro environments of the examined industries are studied. The main methods used are the PESTE-model (macro) and Porter’s five forces model (micro). The study applies a qualitative approach and the data is gathered by interviewing a group of experts from the cement, concrete and construction industries. The result of the paper is an overall trend analysis for the Finnish cement based building sector, which is based on ‘sub trend analyses’ concerning four identified sub-sectors of the Finnish construction industry. The results are a combination of findings from these sub-sectors and the analyzed data that deals with the studied sector’s macro and micro environment. The conclusions provide an overall picture of the examined sectors’ potential future as a whole and by defined sub-sectors of the construction industry. The recognition of future trends in different areas of the construction industry can be applied as a means for an industry actor’s decision making and in estimating the types of construction that are likely to grow or decline. Finally, based on the analyzed data and conclusions, the commissioning company is provided with a brief SWOT analysis, that provides additional tools for decision making and planning processes regarding the future.

Value creation through business model in international entrepreneurship

Business model in the context of international entrepreneurship is a rather new topic in academic literature. The objective of this thesis is to examine value creation through business models in internationally entrepreneurial firms. The study examines value creation through the two partner interfaces and the customer interface of a company. Central for the study is the consideration of also the partners’ incentives. Business model construct is studied by defining the concept, examining its elements and the relationship with strategy – concluding with value creation through the concept. The international entrepreneurship chapter focuses on internationally entrepreneurial firms, inspecting the drivers behind international entrepreneurship and studying value network concept. Value creation functions as a driving theme in the theory discussion. The empirical research of the study focuses on eight Finnish internationally entrepreneurial software companies. The study is conducted as a qualitative cross-case analysis building on the single case company business model analyses. The findings suggest that the business models of software companies incorporate vast similarities. However, the degree of international experience has influence on the companies’ value creation and the way they organize their activities both in upstream and downstream of the value chain.

Developing a Brand Positioning Model for International Strategic Marketing

The target of this thesis is to develop a brand positioning process model for the case company’s international operations. The model will make the process more effective and decrease the risk of relevant aspects being forgotten. The focus is on the international operations although generally the brand positioning can be seen as a standardized subject and, thus, there is no need to distinguish market areas. Constructive research approach is chosen as a research method. Internal interviews are done in order to give the much needed insight about the case company’s current processes and circumstances. Based on theory, interviews as well as internal and external material the model is built. The most difficult part in building the model is to determine the order of each phase. Also, deciding the number of each phase can be problematic. The model should be brief and assertive in order to reduce the risk of misunderstanding between employees from different units. Based on the analysis of the interviews and the theory the brand positioning process model is presented with indication of the order of each phase. The model is divided to three main groups: Analyzing the Environment, Determining the Brand Position, and Documenting the BPS. The benefits of the model are that overlapping work can be reduced, too similar brands can be noticed and it is easier to train new employees.

Improving alumni administration by self-built information system : building IMMIT connected' to enhance alumni connection

Alumni are considered as precious resource of the institutions, thus improving alumni adminis-tration is critical. In information era, alumni administration is assisted by widespread information technology, such as social network sites. This paper aims to discover if a self-built information sys-tem would enhance alumni connection in the IMMIT context, and what kind of attributes would be helpful applying to the special context. The current online alumni services at other universities and at the IMMIT host university are analyzed, and then social media is introduced. After illustrating the social capital existing in IM-MIT, the type of the self-built information system is suggested, following an interpretation of the prototype. Two research models are utilized in this article: TAM and intentional social action model. The second model is adjusted with proposed parameters. Afterwards, a survey and an interview protocol are designed under the guidance of the models. The results are analyzed in several groups, and the proposed parameters are tested. A conclusion is drawn to indicate how to improve alumni‟s intention to use and how to achieve a better-accepted design.

Power cycling lifetime estimation of IGBT power modules based on chip temperature modeling

In this doctoral thesis, methods to estimate the expected power cycling life of power semiconductor modules based on chip temperature modeling are developed. Frequency converters operate under dynamic loads in most electric drives. The varying loads cause thermal expansion and contraction, which stresses the internal boundaries between the material layers in the power module. Eventually, the stress wears out the semiconductor modules. The wear-out cannot be detected by traditional temperature or current measurements inside the frequency converter. Therefore, it is important to develop a method to predict the end of the converter lifetime. The thesis concentrates on power-cycling-related failures of insulated gate bipolar transistors. Two types of power modules are discussed: a direct bonded copper (DBC) sandwich structure with and without a baseplate. Most common failure mechanisms are reviewed, and methods to improve the power cycling lifetime of the power modules are presented. Power cycling curves are determined for a module with a lead-free solder by accelerated power cycling tests. A lifetime model is selected and the parameters are updated based on the power cycling test results. According to the measurements, the factor of improvement in the power cycling lifetime of modern IGBT power modules is greater than 10 during the last decade. Also, it is noticed that a 10 C increase in the chip temperature cycle amplitude decreases the lifetime by 40%. A thermal model for the chip temperature estimation is developed. The model is based on power loss estimation of the chip from the output current of the frequency converter. The model is verified with a purpose-built test equipment, which allows simultaneous measurement and simulation of the chip temperature with an arbitrary load waveform. The measurement system is shown to be convenient for studying the thermal behavior of the chip. It is found that the thermal model has a 5 C accuracy in the temperature estimation. The temperature cycles that the power semiconductor chip has experienced are counted by the rainflow algorithm. The counted cycles are compared with the experimentally verified power cycling curves to estimate the life consumption based on the mission profile of the drive. The methods are validated by the lifetime estimation of a power module in a direct-driven wind turbine. The estimated lifetime of the IGBT power module in a direct-driven wind turbine is 15 000 years, if the turbine is located in south-eastern Finland.

The energy efficiency of chilled beam systems in an office building at different climates

Besides the sustaining of healthy and comfortable indoor climate, the air conditioning system should also achieve for energy efficiency. The target indoor climate can be ob-tained with different systems; this study focuses on comparing the energy efficiency of different air conditioning room unit systems in different climates. The calculations are made with dynamic energy simulation software IDA ICE by comparing the indoor cli-mate and energy consumption of an office building with different systems in different climates. The aim of the study is to compare the energy efficiency of chilled beam systems to other common systems: variable air volume, fan coil and radiant ceiling systems. Besides the annual energy consumption also the sustainability of target indoor climate is compared between the simulations. Another aim is to provide conclusions to be used in the product development of the chilled beam systems’ energy efficiency. The adaptable chilled beam system and the radiant ceiling system prove to be energy efficient independent of the climate. The challenge of reliable comparison is that other systems are not able to reach the target indoor climate as well as the others. The complex calculation environment of the simulation software, made assumptions and excluding of the financial aspects complicate comparing the big picture. The results show that the development of the chilled beam systems should concentrate on energy efficient night heating, flexible demand based ventilation and capacity control and possibilities on integrating the best practices with other systems. 

Reflexive Space. A Constructionist Model of the Russian Reflexive Marker

This study examines the structure of the Russian Reflexive Marker ( ся/-сь) and offers a usage-based model building on Construction Grammar and a probabilistic view of linguistic structure. Traditionally, reflexive verbs are accounted for relative to non-reflexive verbs. These accounts assume that linguistic structures emerge as pairs. Furthermore, these accounts assume directionality where the semantics and structure of a reflexive verb can be derived from the non-reflexive verb. However, this directionality does not necessarily hold diachronically. Additionally, the semantics and the patterns associated with a particular reflexive verb are not always shared with the non-reflexive verb. Thus, a model is proposed that can accommodate the traditional pairs as well as for the possible deviations without postulating different systems. A random sample of 2000 instances marked with the Reflexive Marker was extracted from the Russian National Corpus and the sample used in this study contains 819 unique reflexive verbs. This study moves away from the traditional pair account and introduces the concept of Neighbor Verb. A neighbor verb exists for a reflexive verb if they share the same phonological form excluding the Reflexive Marker. It is claimed here that the Reflexive Marker constitutes a system in Russian and the relation between the reflexive and neighbor verbs constitutes a cross-paradigmatic relation. Furthermore, the relation between the reflexive and the neighbor verb is argued to be of symbolic connectivity rather than directionality. Effectively, the relation holding between particular instantiations can vary. The theoretical basis of the present study builds on this assumption. Several new variables are examined in order to systematically model variability of this symbolic connectivity, specifically the degree and strength of connectivity between items. In usage-based models, the lexicon does not constitute an unstructured list of items. Instead, items are assumed to be interconnected in a network. This interconnectedness is defined as Neighborhood in this study. Additionally, each verb carves its own niche within the Neighborhood and this interconnectedness is modeled through rhyme verbs constituting the degree of connectivity of a particular verb in the lexicon. The second component of the degree of connectivity concerns the status of a particular verb relative to its rhyme verbs. The connectivity within the neighborhood of a particular verb varies and this variability is quantified by using the Levenshtein distance. The second property of the lexical network is the strength of connectivity between items. Frequency of use has been one of the primary variables in functional linguistics used to probe this. In addition, a new variable called Constructional Entropy is introduced in this study building on information theory. It is a quantification of the amount of information carried by a particular reflexive verb in one or more argument constructions. The results of the lexical connectivity indicate that the reflexive verbs have statistically greater neighborhood distances than the neighbor verbs. This distributional property can be used to motivate the traditional observation that the reflexive verbs tend to have idiosyncratic properties. A set of argument constructions, generalizations over usage patterns, are proposed for the reflexive verbs in this study. In addition to the variables associated with the lexical connectivity, a number of variables proposed in the literature are explored and used as predictors in the model. The second part of this study introduces the use of a machine learning algorithm called Random Forests. The performance of the model indicates that it is capable, up to a degree, of disambiguating the proposed argument construction types of the Russian Reflexive Marker. Additionally, a global ranking of the predictors used in the model is offered. Finally, most construction grammars assume that argument construction form a network structure. A new method is proposed that establishes generalization over the argument constructions referred to as Linking Construction. In sum, this study explores the structural properties of the Russian Reflexive Marker and a new model is set forth that can accommodate both the traditional pairs and potential deviations from it in a principled manner.

On state and parameter estimation in chaotic systems

State-of-the-art predictions of atmospheric states rely on large-scale numerical models of chaotic systems. This dissertation studies numerical methods for state and parameter estimation in such systems. The motivation comes from weather and climate models and a methodological perspective is adopted. The dissertation comprises three sections: state estimation, parameter estimation and chemical data assimilation with real atmospheric satellite data. In the state estimation part of this dissertation, a new filtering technique based on a combination of ensemble and variational Kalman filtering approaches, is presented, experimented and discussed. This new filter is developed for large-scale Kalman filtering applications. In the parameter estimation part, three different techniques for parameter estimation in chaotic systems are considered. The methods are studied using the parameterized Lorenz 95 system, which is a benchmark model for data assimilation. In addition, a dilemma related to the uniqueness of weather and climate model closure parameters is discussed. In the data-oriented part of this dissertation, data from the Global Ozone Monitoring by Occultation of Stars (GOMOS) satellite instrument are considered and an alternative algorithm to retrieve atmospheric parameters from the measurements is presented. The validation study presents first global comparisons between two unique satellite-borne datasets of vertical profiles of nitrogen trioxide (NO3), retrieved using GOMOS and Stratospheric Aerosol and Gas Experiment III (SAGE III) satellite instruments. The GOMOS NO3 observations are also considered in a chemical state estimation study in order to retrieve stratospheric temperature profiles. The main result of this dissertation is the consideration of likelihood calculations via Kalman filtering outputs. The concept has previously been used together with stochastic differential equations and in time series analysis. In this work, the concept is applied to chaotic dynamical systems and used together with Markov chain Monte Carlo (MCMC) methods for statistical analysis. In particular, this methodology is advocated for use in numerical weather prediction (NWP) and climate model applications. In addition, the concept is shown to be useful in estimating the filter-specific parameters related, e.g., to model error covariance matrix parameters.

Model based analysis of the post-combustion calcium looping process for carbon dioxide capture

This thesis presents a one-dimensional, semi-empirical dynamic model for the simulation and analysis of a calcium looping process for post-combustion CO2 capture. Reduction of greenhouse emissions from fossil fuel power production requires rapid actions including the development of efficient carbon capture and sequestration technologies. The development of new carbon capture technologies can be expedited by using modelling tools. Techno-economical evaluation of new capture processes can be done quickly and cost-effectively with computational models before building expensive pilot plants. Post-combustion calcium looping is a developing carbon capture process which utilizes fluidized bed technology with lime as a sorbent. The main objective of this work was to analyse the technological feasibility of the calcium looping process at different scales with a computational model. A one-dimensional dynamic model was applied to the calcium looping process, simulating the behaviour of the interconnected circulating fluidized bed reactors. The model incorporates fundamental mass and energy balance solvers to semi-empirical models describing solid behaviour in a circulating fluidized bed and chemical reactions occurring in the calcium loop. In addition, fluidized bed combustion, heat transfer and core-wall layer effects were modelled. The calcium looping model framework was successfully applied to a 30 kWth laboratory scale and a pilot scale unit 1.7 MWth and used to design a conceptual 250 MWth industrial scale unit. Valuable information was gathered from the behaviour of a small scale laboratory device. In addition, the interconnected behaviour of pilot plant reactors and the effect of solid fluidization on the thermal and carbon dioxide balances of the system were analysed. The scale-up study provided practical information on the thermal design of an industrial sized unit, selection of particle size and operability in different load scenarios.

Modeling and Parameter Estimation of Double-Star Permanent Magnet Synchronous Machines

The power rating of wind turbines is constantly increasing; however, keeping the voltage rating at the low-voltage level results in high kilo-ampere currents. An alternative for increasing the power levels without raising the voltage level is provided by multiphase machines. Multiphase machines are used for instance in ship propulsion systems, aerospace applications, electric vehicles, and in other high-power applications including wind energy conversion systems. A machine model in an appropriate reference frame is required in order to design an efficient control for the electric drive. Modeling of multiphase machines poses a challenge because of the mutual couplings between the phases. Mutual couplings degrade the drive performance unless they are properly considered. In certain multiphase machines there is also a problem of high current harmonics, which are easily generated because of the small current path impedance of the harmonic components. However, multiphase machines provide special characteristics compared with the three-phase counterparts: Multiphase machines have a better fault tolerance, and are thus more robust. In addition, the controlled power can be divided among more inverter legs by increasing the number of phases. Moreover, the torque pulsation can be decreased and the harmonic frequency of the torque ripple increased by an appropriate multiphase configuration. By increasing the number of phases it is also possible to obtain more torque per RMS ampere for the same volume, and thus, increase the power density. In this doctoral thesis, a decoupled d–q model of double-star permanent-magnet (PM) synchronous machines is derived based on the inductance matrix diagonalization. The double-star machine is a special type of multiphase machines. Its armature consists of two three-phase winding sets, which are commonly displaced by 30 electrical degrees. In this study, the displacement angle between the sets is considered a parameter. The diagonalization of the inductance matrix results in a simplified model structure, in which the mutual couplings between the reference frames are eliminated. Moreover, the current harmonics are mapped into a reference frame, in which they can be easily controlled. The work also presents methods to determine the machine inductances by a finite-element analysis and by voltage-source inverters on-site. The derived model is validated by experimental results obtained with an example double-star interior PM (IPM) synchronous machine having the sets displaced by 30 electrical degrees. The derived transformation, and consequently, the decoupled d–q machine model, are shown to model the behavior of an actual machine with an acceptable accuracy. Thus, the proposed model is suitable to be used for the model-based control design of electric drives consisting of double-star IPM synchronous machines.