Demographic Modelling of Human Population Growth

This work presents models and methods that have been used in producing forecasts of population growth. The work is intended to emphasize the reliability bounds of the model forecasts. Leslie model and various versions of logistic population models are presented. References to literature and several studies are given. A lot of relevant methodology has been developed in biological sciences. The Leslie modelling approach involves the use of current trends in mortality,fertility, migration and emigration. The model treats population divided in age groups and the model is given as a recursive system. Other group of models is based on straightforward extrapolation of census data. Trajectories of simple exponential growth function and logistic models are used to produce the forecast. The work presents the basics of Leslie type modelling and the logistic models, including multi- parameter logistic functions. The latter model is also analysed from model reliability point of view. Bayesian approach and MCMC method are used to create error bounds of the model predictions.

Eye Fundus Image Analysis for Automatic Detection of Diabetic Retinopathy

Diabetes is a rapidly increasing worldwide problem which is characterised by defective metabolism of glucose that causes long-term dysfunction and failure of various organs. The most common complication of diabetes is diabetic retinopathy (DR), which is one of the primary causes of blindness and visual impairment in adults. The rapid increase of diabetes pushes the limits of the current DR screening capabilities for which the digital imaging of the eye fundus (retinal imaging), and automatic or semi-automatic image analysis algorithms provide a potential solution. In this work, the use of colour in the detection of diabetic retinopathy is statistically studied using a supervised algorithm based on one-class classification and Gaussian mixture model estimation. The presented algorithm distinguishes a certain diabetic lesion type from all other possible objects in eye fundus images by only estimating the probability density function of that certain lesion type. For the training and ground truth estimation, the algorithm combines manual annotations of several experts for which the best practices were experimentally selected. By assessing the algorithm’s performance while conducting experiments with the colour space selection, both illuminance and colour correction, and background class information, the use of colour in the detection of diabetic retinopathy was quantitatively evaluated. Another contribution of this work is the benchmarking framework for eye fundus image analysis algorithms needed for the development of the automatic DR detection algorithms. The benchmarking framework provides guidelines on how to construct a benchmarking database that comprises true patient images, ground truth, and an evaluation protocol. The evaluation is based on the standard receiver operating characteristics analysis and it follows the medical practice in the decision making providing protocols for image- and pixel-based evaluations. During the work, two public medical image databases with ground truth were published: DIARETDB0 and DIARETDB1. The framework, DR databases and the final algorithm, are made public in the web to set the baseline results for automatic detection of diabetic retinopathy. Although deviating from the general context of the thesis, a simple and effective optic disc localisation method is presented. The optic disc localisation is discussed, since normal eye fundus structures are fundamental in the characterisation of DR.

Generative Part-Based Gabor Object Detector

Object detection is a fundamental task of computer vision that is utilized as a core part in a number of industrial and scientific applications, for example, in robotics, where objects need to be correctly detected and localized prior to being grasped and manipulated. Existing object detectors vary in (i) the amount of supervision they need for training, (ii) the type of a learning method adopted (generative or discriminative) and (iii) the amount of spatial information used in the object model (model-free, using no spatial information in the object model, or model-based, with the explicit spatial model of an object). Although some existing methods report good performance in the detection of certain objects, the results tend to be application specific and no universal method has been found that clearly outperforms all others in all areas. This work proposes a novel generative part-based object detector. The generative learning procedure of the developed method allows learning from positive examples only. The detector is based on finding semantically meaningful parts of the object (i.e. a part detector) that can provide additional information to object location, for example, pose. The object class model, i.e. the appearance of the object parts and their spatial variance, constellation, is explicitly modelled in a fully probabilistic manner. The appearance is based on bio-inspired complex-valued Gabor features that are transformed to part probabilities by an unsupervised Gaussian Mixture Model (GMM). The proposed novel randomized GMM enables learning from only a few training examples. The probabilistic spatial model of the part configurations is constructed with a mixture of 2D Gaussians. The appearance of the parts of the object is learned in an object canonical space that removes geometric variations from the part appearance model. Robustness to pose variations is achieved by object pose quantization, which is more efficient than previously used scale and orientation shifts in the Gabor feature space. Performance of the resulting generative object detector is characterized by high recall with low precision, i.e. the generative detector produces large number of false positive detections. Thus a discriminative classifier is used to prune false positive candidate detections produced by the generative detector improving its precision while keeping high recall. Using only a small number of positive examples, the developed object detector performs comparably to state-of-the-art discriminative methods.

Theoretical and numerical study of thermomagnetic convection in magnetic fluids

In this thesis, the magnetic field control of convection instabilities and heat and mass transfer processesin magnetic fluids have been investigated by numerical simulations and theoretical considerations. Simulation models based on finite element and finite volume methods have been developed. In addition to standard conservation equations, themagnetic field inside the simulation domain is calculated from Maxwell equations and the necessary terms to take into account for the magnetic body force and magnetic dissipation have been added to the equations governing the fluid motion.Numerical simulations of magnetic fluid convection near the threshold supportedexperimental observations qualitatively. Near the onset of convection the competitive action of thermal and concentration density gradients leads to mostly spatiotemporally chaotic convection with oscillatory and travelling wave regimes, previously observed in binary mixtures and nematic liquid crystals. In many applications of magnetic fluids, the heat and mass transfer processes including the effects of external magnetic fields are of great importance. In addition to magnetic fluids, the concepts and the simulation models used in this study may be applied also to the studies of convective instabilities in ordinary fluids as well as in other binary mixtures and complex fluids.

Comprehensive Study of Crystal Growth from Solution

Crystal growth is an essential phase in crystallization kinetics. The rate of crystal growth provides significant information for the design and control of crystallization processes; nevertheless, obtaining accurate growth rate data is still challenging due to a number of factors that prevail in crystal growth. In industrial crystallization, crystals are generally grown from multi-componentand multi-particle solutions under complicated hydrodynamic conditions; thus, it is crucial to increase the general understanding of the growth kinetics in these systems. The aim of this work is to develop a model of the crystal growth rate from solution. An extensive literature review of crystal growth focuses on themodelling of growth kinetics and thermodynamics, and new measuring techniques that have been introduced in the field of crystallization. The growth of a singlecrystal is investigated in binary and ternary systems. The binary system consists of potassium dihydrogen phosphate (KDP, crystallizing solute) and water (solvent), and the ternary system includes KDP, water and an organic admixture. The studied admixtures, urea, ethanol and 1-propanol, are employed at relatively highconcentrations (of up to 5.0 molal). The influence of the admixtures on the solution thermodynamics is studied using the Pitzer activity coefficient model. Theprediction method of the ternary solubility in the studied systems is introduced and verified. The growth rate of the KDP (101) face in the studied systems aremeasured in the growth cell as a function of supersaturation, the admixture concentration, the solution velocity over a crystal and temperature. In addition, the surface morphology of the KDP (101) face is studied using ex situ atomic force microscopy (AFM). The crystal growth rate in the ternary systems is modelled on the basis of the two-step growth model that contains the Maxwell-Stefan (MS) equations and a surface-reaction model. This model is used together with measuredcrystal growth rate data to develop a new method for the evaluation of the model parameters. The validation of the model is justified with experiments. The crystal growth rate in an imperfectly mixed suspension crystallizer is investigatedusing computational fluid dynamics (CFD). A solid-liquid suspension flow that includes multi-sized particles is described by the multi-fluid model as well as by a standard k-epsilon turbulence model and an interface momentum transfer model. The local crystal growth rate is determined from calculated flow information in a diffusion-controlled crystal growth regime. The calculated results are evaluated experimentally.

Tyypin 1 diabeteksen matemaattisista ennustamismenetelmistä

Tutkielmassa käsitellään matemaattisia ennustamismenetelmiä, jotka soveltuvat tyypin 1 diabeteksen ennustamiseen. Aluksi esitellään menetelmiä, jotka soveltuvat puuttuvia havaintoja sisältävien aineistojen paikkaamiseen. Paikattua aineistoa on mahdollista analysoida useilla tavallisilla tilastollisilla menetelmillä, jotka sopivat täydellisiin aineistoihin. Seuraavaksi pyritään mallintamaan aineistoa semiparametrisilla komponenttimalleilla (eng. mixture model), jolloin mallin muotoa ei ole tiukasti etukäteen rajoitettu. Sen jälkeen sovelletaan kolmea luokittelevaa ennustajaa: logistista regressiomallia, eteenpäinsyöttävää yhden piilotason neuroverkkoa ja SVM-menetelmää (eng. support vector machine). Esiteltäviä menetelmiä on sovellettu todelliseen aineistoon, joka on kerätty Turun yliopistossa käynnissä olevassa tutkimusprojektissa. Projektin tavoitteena on oppia ennustamaan ja ehkäisemään tyypin 1 diabetesta (Type 1 diabetes prediction and prevention project, lyh. DIPP-projekti). Erityisesti projektissa on pyritty löytämään uusia tuntemattomia taudinaiheuttajia. Tässä tutkielmassa paneudutaan sen sijaan kerätyn havaintoaineiston matemaattisiin analysointimenetelmiin. Parhaat ennusteet saatiin perinteisellä logistisella regressiomallilla. Tutkielmassa kuitenkin todetaan, että tulevaisuudessa on mahdollista löytää parempia ennustajia parantamalla muita edellä mainittuja menetelmiä. Erityisesti SVM-menetelmä ansaitsisi lisähuomiota, sillä tässä tutkielmassa sitä sovellettiin vain kaikkein yksinkertaisimmassa muodossa.

Hiljaisen tiedon siirtämisen prosessi mentoroinnissa – näkökulmana kognitiivinen oppipoikamalli

Hiljainen tieto muodostaa organisaatioiden keskeisen kilpailutekijän, sillä sitä on vaikea kopioida. Hiljaista tietoa pyritään siirtämään erilaisia osaamisen kehittämisen menetelmiä hyödyntäen. Tässä tutkielmassa tutkitaan, miten hiljaista tietoa siirretään mentoroinnissa. Mentorointiin liittyvissä tutkimuksissa ei ole tutkittu sitä vuorovaikutukseen perustuvaa prosessia, jonka aikana hiljaista tietoa siirretään mentorilta aktorille. Tämä tutkielma toi lisää tietoa tähän tutkimusaukkoon. Tutkielman teoreettisessa osiossa esiteltiin kolme näkökulmaa, jotka muo-dostivat tutkielman viitekehyksen: hiljainen tieto ja sen siirtäminen, mentorointi sekä kognitiivinen oppipoikamalli. Tutkimusmenetelmänä käytettiin fenomenografista tapaustutkimusta. Tutkimuksen kohderyhmän muodostivat neljä mentori-aktori –paria, joita haastateltiin teemahaastattelulla. Empiiriset tulokset osoittivat, että hiljaisen tiedon siirtäminen mentoroinnissa tapahtui kognitiivisen oppipoikamallin vaiheita hyödyntäen. Kaikki kognitiivisen oppipoikamallin vaiheet esiintyivät mentorointiprosessissa. Siirrettävässä hiljaisessa tiedossa näyttäytyivät tiedon toiminnallinen, situationaalinen ja sosiaalinen luonne. Keskeisimmiksi hiljaisen tiedon siirtämisen menetelmiksi osoittautuivat mentorin läsnäolo, kuuntelu, kysymysten tekeminen ja aktorin oivalluttaminen. Tutkielman keskeisenä tuloksena ja toimenpide-ehdotuksena esitettiin hiljaisen tiedon siirtämisen malli mentoroinnissa, joka kehitettiin tutkimuksen teoreettisen viitekehyksen ja tutkimuksesta saatujen tulosten pohjalta.

Differentiation of Metabolically Distinct Areas within Head and Neck Region using Dynamic 18F-FDG Positron Emission Tomography Imaging

Positron Emission Tomography (PET) using ¹⁸F-FDG is playing a vital role in the diagnosis and treatment planning of cancer. However, the most widely used radiotracer, ¹⁸F-FDG, is not specific for tumours and can also accumulate in inflammatory lesions as well as normal physiologically active tissues making diagnosis and treatment planning complicated for the physicians. Malignant, inflammatory and normal tissues are known to have different pathways for glucose metabolism which could possibly be evident from different characteristics of the time activity curves from a dynamic PET acquisition protocol. Therefore, we aimed to develop new image analysis methods, for PET scans of the head and neck region, which could differentiate between inflammation, tumour and normal tissues using this functional information within these radiotracer uptake areas. We developed different dynamic features from the time activity curves of voxels in these areas and compared them with the widely used static parameter, SUV, using Gaussian Mixture Model algorithm as well as K-means algorithm in order to assess their effectiveness in discriminating metabolically different areas. Moreover, we also correlated dynamic features with other clinical metrics obtained independently of PET imaging. The results show that some of the developed features can prove to be useful in differentiating tumour tissues from inflammatory regions and some dynamic features also provide positive correlations with clinical metrics. If these proposed methods are further explored then they can prove to be useful in reducing false positive tumour detections and developing real world applications for tumour diagnosis and contouring.

Statistical segmentation methods and color variance analysis of retinal images

In this research, the effectiveness of Naive Bayes and Gaussian Mixture Models classifiers on segmenting exudates in retinal images is studied and the results are evaluated with metrics commonly used in medical imaging. Also, a color variation analysis of retinal images is carried out to find how effectively can retinal images be segmented using only the color information of the pixels.

Suitability of LiDAR technology for forest inventory in Russia

LiDAR is an advanced remote sensing technology with many applications, including forest inventory. The most common type is ALS (airborne laser scanning). The method is successfully utilized in many developed markets, where it is replacing traditional forest inventory methods. However, it is innovative for Russian market, where traditional field inventory dominates. ArboLiDAR is a forest inventory solution that engages LiDAR, color infrared imagery, GPS ground control plots and field sample plots, developed by Arbonaut Ltd. This study is an industrial market research for LiDAR technology in Russia focused on customer needs. Russian forestry market is very attractive, because of large growing stock volumes. It underwent drastic changes in 2006, but it is still in transitional stage. There are several types of forest inventory, both with public and private funding. Private forestry enterprises basically need forest inventory in two cases – while making coupe demarcation before timber harvesting and as a part of forest management planning, that is supposed to be done every ten years on the whole leased territory. The study covered 14 companies in total that include private forestry companies with timber harvesting activities, private forest inventory providers, state subordinate companies and forestry software developer. The research strategy is multiple case studies with semi-structured interviews as the main data collection technique. The study focuses on North-West Russia, as it is the most developed Russian region in forestry. The research applies the Voice of the Customer (VOC) concept to elicit customer needs of Russian forestry actors and discovers how these needs are met. It studies forest inventory methods currently applied in Russia and proposes the model of method comparison, based on Multi-criteria decision making (MCDM) approach, mainly on Analytical Hierarchy Process (AHP). Required product attributes are classified in accordance with Kano model. The answer about suitability of LiDAR technology is ambiguous, since many details should be taken into account.

‘Théorie’, ‘hypothèse’, ‘modèle’ et ‘méthode’ dans le domaine de la linguistique: perspectives sémasiologique et onomasiologique

'Theory', 'hypothesis', 'model' and 'method' in linguistics: Semasiological and onomasiological perspectives The subject of this thesis is the use of generic scientific terms, in particular the four terms 'theory', 'hypothesis', 'model' and 'method', in linguistic research articles written in French and in Finnish. The thesis examines the types of scientific constructs to which these terms are applied, and seeks to explain the variation in the use of each term. A second objective of the thesis is to analyze the relationships among these terms, and the factors determining the choices made by writers. With its focus on the authentic use of generic scientific terms, the thesis complements the normative and theoretical descriptions of these terms in Science Studies and offers new information on actual writing practices. This thesis adheres to functional and usage-based linguistics, drawing its theoretical background from cognitive linguistics and from functional approaches to terminology. The research material consisted of 120 research articles (856 569 words), representing different domains of linguistics and written in French or Finnish (60 articles in each language). The articles were extracted from peer-reviewed scientific journals and were published between 2000 and 2010. The use of generic scientific terms in the material has been examined from semasiological and onomasiological perspectives. In the first stage, different usages related to each of the four central terms were analyzed. In the second stage, the analysis was extended to other terms and expressions, such as 'theoretical framework', 'approach' and ‘claim’, which were used to name scientific constructs similar to the four terms analyzed in the first stage. Finally, in order to account for the writer’s choice among the terms, a mixed methods approach was adopted, based on the results of a previously conducted questionnaire concerning the differences between these terms as experienced by linguists themselves. Despite the general ideal that scientific terms should be carefully defined, the study shows that the use of these central terms is not without ambiguity. What is understood by these terms may vary according to different conceptual and stylistic factors as well as epistemic and disciplinary traditions. In addition to their polysemy, the semantic potentials of these terms are in part overlapping. In most cases, the variation in the use of these terms is not likely to cause serious misunderstanding. Rather, it allows the researcher to express a specific conceptualization of the scientific constructs mentioned in the article. The discipline of linguistics, however, would benefit from a more elaborate metatheoretical discussion.

Bayesian Estimation of Noise

In mathematical modeling the estimation of the model parameters is one of the most common problems. The goal is to seek parameters that fit to the measurements as well as possible. There is always error in the measurements which implies uncertainty to the model estimates. In Bayesian statistics all the unknown quantities are presented as probability distributions. If there is knowledge about parameters beforehand, it can be formulated as a prior distribution. The Bays’ rule combines the prior and the measurements to posterior distribution. Mathematical models are typically nonlinear, to produce statistics for them requires efficient sampling algorithms. In this thesis both Metropolis-Hastings (MH), Adaptive Metropolis (AM) algorithms and Gibbs sampling are introduced. In the thesis different ways to present prior distributions are introduced. The main issue is in the measurement error estimation and how to obtain prior knowledge for variance or covariance. Variance and covariance sampling is combined with the algorithms above. The examples of the hyperprior models are applied to estimation of model parameters and error in an outlier case.

A generalized 1D particle transport method for convection diffusion reaction model

This work is devoted to the development of numerical method to deal with convection diffusion dominated problem with reaction term, non - stiff chemical reaction and stiff chemical reaction. The technique is based on the unifying Eulerian - Lagrangian schemes (particle transport method) under the framework of operator splitting method. In the computational domain, the particle set is assigned to solve the convection reaction subproblem along the characteristic curves created by convective velocity. At each time step, convection, diffusion and reaction terms are solved separately by assuming that, each phenomenon occurs separately in a sequential fashion. Moreover, adaptivities and projection techniques are used to add particles in the regions of high gradients (steep fronts) and discontinuities and transfer a solution from particle set onto grid point respectively. The numerical results show that, the particle transport method has improved the solutions of CDR problems. Nevertheless, the method is time consumer when compared with other classical technique e.g., method of lines. Apart from this advantage, the particle transport method can be used to simulate problems that involve movingsteep/smooth fronts such as separation of two or more elements in the system.

Solving Leontief Input-Output Model with Trapezoidal Fuzzy Numbers and Gauss-Seidel Algorithm

In this work a fuzzy linear system is used to solve Leontief input-output model with fuzzy entries. For solving this model, we assume that the consumption matrix from di erent sectors of the economy and demand are known. These assumptions heavily depend on the information obtained from the industries. Hence uncertainties are involved in this information. The aim of this work is to model these uncertainties and to address them by fuzzy entries such as fuzzy numbers and LR-type fuzzy numbers (triangular and trapezoidal). Fuzzy linear system has been developed using fuzzy data and it is solved using Gauss-Seidel algorithm. Numerical examples show the e ciency of this algorithm. The famous example from Prof. Leontief, where he solved the production levels for U.S. economy in 1958, is also further analyzed.