Browsing by Title

In little over a decade, cryptocurrencies have become a highly speculative asset class in global financial markets, with Bitcoin leading the way. Throughout its relatively brief history, the price of bitcoin has gone through multiple cycles of growth and decline. As a consequence, Bitcoin has become a widely discussed – and polarizing – topic on Twitter. This work studies whether the sentiment of popular Bitcoin-related tweets can be used to predict the future price movements of bitcoin. In total, seven different algorithms are evaluated: Vector Autoregression, Vector Autoregression Moving-Average, Random Forest, XGBoost, LightGBM, Long Short-Term Memory, and Gated Recurrent Unit. By applying lexicon-based sentiment analysis, and heuristic filtering of tweets, it was discovered that sentiment-based features of popular tweets improve the prediction accuracy over baseline features (open-high-low-close data) in five of the seven algorithms tested. The tree-based algorithms (Random Forest, XGBoost, LightGBM) generally had the lowest prediction errors, while the neural network algorithms (Light Short-Term Memory and Gated Recurrent Unit) had the poorest performance. The findings suggest that the sentiment of popular Bitcoin-related tweets can be an important feature in predicting the future price movements of bitcoin.

Bierol, tetramer lignin model compound was synthesized from guaicylglycerol-ß-guaiacyl ether (Erol) by oxidative dimerization using Horseradish peroxidase (HRP) / H2O2 biomimetic oxidation system. Starting material Erol, a (β-O-4) type lignin model compound was synthesized in 80% overall yield in six steps from acetovanillone. The reactions progresses were associated with thin layer chromatography (TLC) and the purity of each product was obtained by High Performance Liquid Chromatography (HPLC) analysis techniques. The 1H and 13C NMR characteristics of the resulting products were studied using 1H and 13C NMR spectra as well as 2D NMR, in particular (Heteronuclear Single Quantum Coherence) HSQC. The objectives of this series of investigation was three-fold: (a) to synthesize dimeric lignin model compounds (ß-O-4), (b) to synthesize tetrameric lignin model compounds, (c) to study 1H and 13C NMR and 2D NMR spectroscopic characteristics of compounds. The research conducted in this project was successful as the model compounds were synthesized and characterized. Accordingly, the results were discussed in terms of signal assignments that confirmed the molecules structures. The general knowledge of lignin, e.g. chemical structure of lignin, the synthetic methods for synthesizing (ß-O-4) type model compounds, the different methods for lignin oxidative degradation and the various methods of lignin analysis were reviewed.

Problems which ask to compute an optimal solution to its instances are called optimization problems. The maximum satisfiability (MaxSAT) problem is a well-studied combinatorial optimization problem with many applications in domains such as cancer therapy design, electronic markets, hardware debugging and routing. Many problems, including the aforementioned ones, can be encoded in MaxSAT. Thus MaxSAT serves as a general optimization paradigm and therefore advances in MaxSAT algorithms translate to advances in solving other problems. In this thesis, we analyze the effects of MaxSAT preprocessing, the process of reformulating the input instance prior to solving, on the perceived costs of solutions during search. We show that after preprocessing most MaxSAT solvers may misinterpret the costs of non-optimal solutions. Many MaxSAT algorithms use the found non-optimal solutions in guiding the search for solutions and so the misinterpretation of costs may misguide the search. Towards remedying this issue, we introduce and study the concept of locally minimal solutions. We show that for some of the central preprocessing techniques for MaxSAT, the perceived cost of a locally minimal solution to a preprocessed instance equals the cost of the corresponding reconstructed solution to the original instance. We develop a stochastic local search algorithm for MaxSAT, called LMS-SLS, that is prepended with a preprocessor and that searches over locally minimal solutions. We implement LMS-SLS and analyze the performance of its different components, particularly the effects of preprocessing and computing locally minimal solutions, and also compare LMS-SLS with the state-of-the-art SLS solver SATLike for MaxSAT.p

This thesis attempts to show the advantages of Monte Carlo method in pricing and hedging exotic options. The popularity of exotic options increased recently, mostly due to their almost unlimited flexibility and adaptability to any circumstances. The negative side of exotic options is their complexity. Due to that many exotic options does not have analytic solutions. As a result numerical solutions are a necessity. The Monte Carlo method of simulations is very common method in computational finance. Monte Carlo method is based on the analogy between probability and volume. Starting point in pricing and hedging options with Monte Carlo method is stochastic differential equation based on Brownian motion in the Black-Scholes world. The fair option value in the Black-Scholes world is the present value of the expected payoff at expiry under a risk-neutral assumptions. The analysis start from the case of the simple European options and continue with introducing different kinds of exotic options. The dynamic hedging idea is used to derive the Black-Scholes Partial Differential Equation. The numerical approximation of the stochastic differential equation is derived through the lognormal asset price model. The Monte Carlo algorithms are constructed for pricing and delta hedging and then implemented to MATLAB. For generating Monte Carlo simulations is used N(0,1) pseudo-random generator. The analysis is limited to the cases of simple Barrier options, which are one of the most known and used type of the exotic options. Barrier options are path dependent options, which implies that the payoff depends on the path followed by the price of the underlying asset, meaning that barrier options prices are especially sensitive to volatility. That is why, we also introduce the variance reduction techniques by antithetic variates. For hedging barrier options were chosen the dynamic delta-hedging and static hedging strategies. All calculations and figures in the examples were made in MATLAB.

This study focuses on looking at different factors that have been shown to explain why certain farms are vulnerable to crop-raiding by primates. The type of wildlife value orientations present among the study sample are also looked into in this study. Furthermore, how these wildlife value orientations affect how the community members perceive and react to human-wildlife conflicts is also discussed. The data collection was conducted in May and June 2015 around Ngangao indigenous cloud forest in the Taita Hills, southeast Kenya. The study is both qualitative and quantitative. Semi-structured household-based questionnaires were used in the primary data collection. Also, 11 local experts were interviewed and two workshops were held where the participants were given topics about wildlife issues and solutions to discuss about. Statistical analysis as well as spatial analysis using GIS were performed in this study. The findings of the study are that the closer a farm is to the forest boundary and the less neighbouring farms there are between the farm and the forest, the more vulnerable that farm is to crop-raiding by Sykes' monkeys. The study could not prove that a specific type of food crop grown in a farm or the type of land use between the farmland and the forest boundary is explaining vulnerability to crop-raiding by primates. Moreover, strong determinants that explain the vulnerability of a certain farm to crop-raiding by vervet monkeys, yellow baboons or bush-babies were not found in this study. The majority of the studied households practice subsistence farming as their main livelihood. Therefore, crop-raiding by wildlife, such as primates, is a severe threat to the food security and livelihoods of local households. The study points out that a majority of the study participants perceive wildlife in a materialistic way, either as threats or as benefits. A smaller share of respondents represent a harmonius wildlife value orientation. Because majority of the local community is likely to represent similar wildlife value orientations, crop-raiding by primates is perceived as a significant problem. Some community members are believed to have reacted to human-primate conflicts by deliberately setting forest fires in the forests of the Taita Hills to deter the marauding primates away. This study suggests that community representatives, local wildlife management and officials should collectively discuss and address the issue of human-primate conflict in the indigenous forests of the Taita Hills so that the community level perceptions are not disregarded in wildlife management. Possible solutions to human-primate conflicts are, for example, providing the most vulnerable households with compensation, incentives or food relief and tools to practice alternative livelihoods. Moreover, relocation of certain groups of primates and planting wild fruit trees inside indigenous forests could ease the problem. Community wildlife association could work as a platform to address the issue. Additionally, sensitizing the local communities about local nature and addressing the awareness gaps regarding problem wildlife reporting could increase the tolerance to wildlife damages at community level and also support the coexistence of humans and wildlife.

The primordial perturbations created by inflation in the early Universe are known to be able to produce significant amount of primordial black holes and gravitational waves with large amplitudes in some inflationary models. Primordial black holes are produced by primordial scalar perturbations and gravitational waves are partly primordial tensor perturbations and partly produced by scalar perturbations. In this thesis we review some of the current literature on the subject and discuss a few inflationary models that are capable of producing primordial scalar perturbations large enough to create a significant amount of primordial black holes. The main focus is on ultra-slow roll inflation with a concrete example potential illustrating the dynamics of the scenario followed by a briefer treatment of some of the alternative models. We start by explaining the necessary background theory for the understanding of the subject at hand. Then we move on to the inflationary models covered in this thesis. After that we explain the production of the primordial black holes and gravitational waves from scalar perturbations. Then we consider primordial black holes as a dark matter candidate and go through the most significant known restrictions on the existence of primordial black holes with different masses. We discuss some of the possible future constraints for the remaining possible mass window for which primordial black holes could explain all of dark matter. We then briefly discuss two planned space-based gravitational wave detectors that may be able to detect gravitational waves created by inflation.

I slutet av 1700-talet gissade Gauss och Legendre att lim_{x→∞} \frac{π (x) log (x)}{x} =1, där π (x) är antalet primtal som är mindre än eller lika med x. Hadamard och de la Vall'ee Poussin bevisade påstående oberoende av varandra år 1896 och resultatet kallas numera primtalssatsen. Efter detta har satsen bevisats på både elementära sätt (Selberg & Erdõs, 1949) och med hjälp av komplexanalys (Newman, 1980). I denna avhandling kommer vi att presentera ett analytiskt bevis av PTS. I beviset kommer vi att utnyttja Riemanns zetafunktion och dess egenskaper. I kapitel 3 behandlar vi komplexanalys. Vi diskuterar bl.a. Eulers gammafunktion, Riemanns zetafunktion och Dirichletserier. Det är en naturlig fortsättning till PTS att diskutera zetafunktionens nollställen, och därför bevisar vi Hardys sats. Efter detta kommer vi att presentera den ökända Riemannhypotesen. Gauss gissade att den logaritmiska integralen \li (x) :=∈t_2^x \frac{dt}{log t} skulle approximera π (x) mycket bra. Vi definierar den s.k. restfunktionen som r(x):=π (x) - \li (x). Vi avslutar avhandlingen genom att bevisa att om det för varje \epsilon >0 gäller att |r(x)|<x^{\frac{1}{2} +\epsilon} så är Riemannhypotesen sann.

This thesis focuses on privacy-preserving statistical inference. We use a probabilistic point of view of privacy called differential privacy. Differential privacy ensures that replacing one individual from the dataset with another individual does not affect the results drastically. There are different versions of the differential privacy. This thesis considers the ε-differential privacy also known as the pure differential privacy, and also a relaxation known as the (ε, δ)-differential privacy. We state several important definitions and theorems of DP. The proofs for most of the theorems are given in this thesis. Our goal is to build a general framework for privacy preserving posterior inference. To achieve this we use an approximative approach for posterior inference called variational Bayesian (VB) methods. We build the basic concepts of variational inference with certain detail and show examples on how to apply variational inference. After giving the prerequisites on both DP and VB we state our main result, the differentially private variational inference (DPVI) method. We use a recently proposed doubly stochastic variational inference (DSVI) combined with Gaussian mechanism to build a privacy-preserving method for posterior inference. We give the algorithm definition and explain its parameters. The DPVI method is compared against the state-of-the-art method for DP posterior inference called the differentially private stochastic gradient Langevin dynamics (DP-SGLD). We compare the performance on two different models, the logistic regression model and the Gaussian mixture model. The DPVI method outperforms DP-SGLD in both tasks.

Over the last years, Location-Based Services (LBSs) have become popular due to the global use of smartphones and improvement in Global Positioning System (GPS) and other positioning methods. Location-based services employ users' location to offer relevant information to users or provide them with useful recommendations. Meanwhile, with the development of social applications, location-based social networking services (LBSNS) have attracted millions of users because the geographic position of users can be used to enhance the services provided by those social applications. Proximity detection, as one type of location-based function, makes LBSNS more flexible and notifies mobile users when they are in proximity. Despite all the desirable features that such applications provide, disclosing the exact location of individuals to a centralized server and/or their social friends might put users at risk of falling their information in wrong hands, since locations may disclose sensitive information about people including political and religious affiliations, lifestyle, health status, etc. Consequently, users might be unwilling to participate in such applications. To this end, private proximity detection schemes enable two parties to check whether they are in close proximity while keeping their exact locations secret. In particular, running a private proximity detection protocol between two parties only results in a boolean value to the querier. Besides, it guarantees that no other information can be leaked to the participants regarding the other party's location. However, most proposed private proximity detection protocols enable users to choose only a simple geometric range on the map, such as a circle or a rectangle, in order to test for proximity. In this thesis, we take inspiration from the field of Computational Geometry and develop two privacy-preserving proximity detection protocols that allow a mobile user to specify an arbitrary complex polygon on the map and check whether his/her friends are located therein. We also analyzed the efficiency of our solutions in terms of computational and communication costs. Our evaluation shows that compared to the similar earlier work, the proposed solution increases the computational efficiency by up to 50%, and reduces the communication overhead by up to 90%. Therefore, we have achieved a significant reduction of computational and communication complexity.

Tässä tutkielmassa tutustutaan proäärellisiin ryhmiin, siis topologisiin ryhmiin, jotka ovat isomorfisia äärellisten topologisten ryhmien muodostaman inverssisysteemin rajan kanssa. Tutkielman alussa esitetään topologisten ryhmien yleistä teoriaa, sekä tutustutaan inverssisysteemeihin yleisesti esittämällä kokoelma näiden ominaisuuksia. Tämän jälkeen tutkielmassa siirrytään käsittelemään proäärellisiä ryhmiä ja ja esitetään tärkeä karakterisaatio, jonka mukaan topologinen ryhmä on proäärellinen jos ja vain jos se on kompakti ja täysin epäyhtenäinen. Tästä seuraa Blairen lauseen nojalla, että jokainen proäärellinen ryhmä on Blairen avaruus. Tutkielman lopuksi käsitellään proäärellisiä täydellistymiä, erityisesti kokonaislukujen pro-p täydellistymämme, siis p-adisille luvuille, annetaan oma kappaleensa, jossa esitetään näiden konstruktio inverssisysteemin rajana sekä äärettömän pitkinä luonnollisina lukuina, sekä esitetään näiden ominaisuuksia. Näistä tärkeimpänä esitetään Henselin lemma, jonka avulla p-adisten lukujen polynomille löydetään juuri, kunhan sille on annettu riittähän hyvä arvio. Tällä tuloksella on käyttöä myös modulaariaritmeriikassa. Viimeisenä tutkielmassa esitetään kokonaislukujen proäärellinen täydellistymä