The cyanide-degrading bacterium Klebsiella oxytoca SYSU-011 was isolated from the waste water of a metal-plating plant. In this study, we found out that K. oxytoca was capable of utilizing tetracyanonickelate {K2[Ni(CN)4]}(TCN) as its sole nitrogen source. This organism could degrade TCN both aerobically (D.O.¡×100¢H) and anaerobically (D.O.¡×0¢H).The addition of ammonia (5 mM) in the growth medium would inhibit TCN-degrading. The TCN-degrading by-product, a greenish precipitate, was found in the spent medium and was identified as nickel cyanide [Ni(CN)2] by FT-IR spectroscopic studies. Ammonia was demonstrated as a product of the TCN-degrading process by K. oxytoca resting cells. The addition of glucose could greatly enhance the TCN-degradation. Nitrogenase was found to be the cyanide degrading enzyme in this organism. The activity of nitrogenase was inhibited by ammonia but could be induced by the addition of TCN or KCN.

Les Andes de Mérida (Venezuela) est caractérisé par la présence de morphologie glaciaire bien préservée entre 2400-4978 m. Les études de ces morphologies glaciares font la contribution pour mieux comprendre et reconstruire la Dernière Glaciation aux Andes Vénézuéliennes. La reconstruction de la Dernière Glaciation (connue dans la région comme Glaciation Mérida) était limitée par la disponibilité de chronologie glaciaire. Cette thèse fait sa contribution à la reconstruction de la Dernière Glaciation et à la connaissance du paléoclimat depuis le Pléistocène Superieur. La méthodologie a involucré l´analyse géomorphologique et l´étude géochronologique. La morphologie glaciaire a été datée avec la technique de nucléides cosmogéniques produits in-situ (10Be). Cette technique est appropriée pour dater les matériels riches en quartz et la période d´intérêt. Les inférences des conditions paleoclimatiques ont été faites sur la base des analyses de la paléo ELA.Des différentes dynamiques glaciares ont été identifiées aux Andes centrales de Mérida. Dans les vallées de Mucubají et Mucuchache, la dynamique a été caractérisée par plusieurs étapes de avancé-arrêt des glaciaires dans un recul général de la glace. Dans les vallées de Gavidia et Mifafí, la disparition des glaciares a été très rapide avec la vitesse de recul plus haute (entre 4-7 km/ky). Divers caractéristiques morpho métriques comme la pente du fond de la vallée, la topographie de la zone d´accumulation (cirques glaciares avec les murs très incliné), des surfaces de la zone d´accumulation et son orientation ont aussi contrôlé les différents dynamiques glaciares.Les avancées des glaciares ont été identifiées entre 2500-4200 m. Les avancées glaciares pendant le MIS 3 ont été reconnues à Sierra Nevada. Ces avancées ont été liées aux hautes précipitations dans le nord de l´Amérique de Sud produites par l´activité de la Zone de Convergence Intertropical. Ces avancées ont été aussi reliées aux conditions plus chaudes et humides dans l´Hémisphère Nord (aussi ces conditions ont été identifiées aux Andes vénézuéliennes et sont connues comme l´Interstadío El Pedregal). Les avancées glaciares pendant le Dernière Maximum Glaciares ont été enregistrées en Sierra Nevada á Mucubají et Las Tapias entre 3100-3600 m. Principalement, les avancées glaciares MIS 2 sont arrivées pendant l´Oldest Dryas-El Caballo Estadío à 17 ka. Ces avancées glaciares ont été reliées aux températures froides dans l´Hémisphère Nord et les températures plus froides enregistrées dans les carottes de glaciares tropicaux.Mots clésDatation par cosmogéniques produits in-situ, morphologie glaciaire, Pléistocène, Dernière Glaciation, LGM, paléo ELA. / The central Mérida Andes (Venezuela) landscape is characterized by the presence of well-preserved glacial landforms located between 2400 and 4978 m a.s.l. Geomorphological studies of these glacial landforms significantly contribute to the Venezuelan Andes glaciations reconstructions. However, Last Glaciation (locally called Mérida Glaciation) was poorly reconstructed because of limited chronological data. This dissertation attempts to contribute to the Last Glaciation reconstruction and paleoclimate knowledge since the late Pleistocene. Accordance this necessity, the methodology involved geomorphological analysis and geochronological study. Glacial landforms were dated based on the Terrestrial Cosmogenic Nuclide dating (10Be). This method is suitable for date quartz-rich materials and for the period of interest. To deduce paleoclimate conditions an analysis based on paleo ELA was developed.In the central Mérida Andes different Late Pleistocene glacier dynamics were identified. In the Mucubají and the Mucuchache valleys, successive stages of glacier stop-advance were identified during an overall glacier withdrawal. In the Gavidia and Mifafí valleys, glacier withdrawal was rapid with the highest retreat rates (between 4-7 km/ky). Morphometric features as glaciers bottom valley slopes, accumulation zone topography (glaciers cirques with steep walls), areas and orientation controlled different glaciers dynamics.Glacier advances were evidenced between 2500-4200 m. MIS 3 glaciers advances has been recognized in the Sierra Nevada. These were related to the highest runoff in the north of South America produced by the Intertropical Convergence Zone (ITCZ) and the local warm and wet climate conditions (locally named El Pedregal Interstadial). LGM glacier advances were recorded in Sierra Nevada in the Mucubají and Las Tapias between 3100-3600 m. MIS 2 Glaciers advances mainly occurred during the Oldest Dryas- El Caballo Stadial at around 17 ka. These glaciers advances correlate to the cold temperatures in the North Hemisphere and the coldest temperatures recorded in tropical ice cores.KeywordsTerrestrial cosmogenic nuclides dating, TCN, cosmogenic dating, glacial landforms, Andes Mérida, Venezuela. Pleistocene, Last Glaciation, LGM, paleo ELA, tropic paleoclimate.

Machine learning models has gained traction as an effective tool for short-term electricity price forecasting, namely day ahead and hourly price forecasting. Efficient and accurate forecasting is crucial for demand and capacity planning to ensure stability and optimal use of resources. This project applies two proven machine learning models, LSTM and TCN, to electricity futures contracts in the Swedish pricing areas SE1 and SE3. Future contracts are used to secure the price of electricity in the future. A multivariate time series of fundamental data that correlates with electricity prices is used as input for the forecasting. Fur- thermore, a portfolio approach for hedging is evaluated based on the predictive performance of the models. The forecasting accuracy of the multivariate TCN model outperform the LSTM model. The optimal hedging strategy based on the TCN model indicated potential cost savings of 1.43% compared to a benchmark method. / Maskininlärnings modeller har vunnit mark som effektiva verktyg för att prognosticera kortsiktiga elpriser, för dagen före och timpriser. Effektiv och korrekt prognosticering är viktigt för att skatta behovs- och kapacitetsplanering för optimal resursanvändning. Det här projektet applicerar två välbeprövade modeller, LSTM och TCN, för att prognosticera terminskontrakt i de två svenska pris- områdena SE1 och SE3. Terminskontrakt används för att säkra elpriser i framtiden. En tidsserie, med flera variabler av fundamental data som korrelerar med elpriser, används för att prognosticera elpriser. Vidare utvärderas en portfölj approach för prissäkring baserat på prognoserna från modellerna. TCN modellen gav högre noggrannhet än LSTM modellen. Optimal prissäkringsstrategi baserad på TCN modeller resulterade i 1.43% lägre elpriser jämfört med bench- marks.

Good generative models for short time series data exist and have been applied for both data augmentation and privacy protection purposes in the past. A common theme for existing generative models is that they all use a recurrent neural network (RNN) architecture, which makes the models limited regarding the length of the sequences. In real world problems, we might have to deal with data containing longer sequences, and it is such data we in this thesis attempt to synthesize. By combining the recently successful TimeGAN framework with a temporal convolutional network component architecture, we generate synthetic sequential data for two toy data sets: sequential MNIST and multivariate sine waves. The results strongly indicate, although relying solely on a visual inspection, that the model manage to capture long temporal dynamics over time and also relations between different features for the multivariate sine waves data set. In order to make our model applicable for real world data sets, we suggest two improvements. Firstly, the validation of the generated data should not only rely on visual inspection, but also ensure that the synthetic data has the same statistical distribution. Secondly, depending on the task, model refinements such that the synthetic samples look even more realistic should be made.

Under åren har intensiv forskning pågått för att förbättra maskininlärningsmodellers förmåga att förutse marknadsrörelser. Trots detta har det, under finanshistorien, inträffat flera händelser, såsom "Flash-crash", som har påverkat marknaden och haft dramatiska konsekvenser för prisrörelserna. Därför är det viktigt att undersöka hur modellerna påverkas av manipulativa handlingar på finansmarknaden för att säkerställa deras robusthet och tillförlitlighet i sådana situationer.  För att genomföra detta arbete har processen delats upp i tre steg. Först har en undersökning av tidigare arbeten gjorts för att identifiera de mest robusta modellerna inom området. Detta gjordes genom att träna modellerna på FI-2010 datasetet, som är ett offentligt tillgängligt dataset för högfrekvent handel med aktier på NASDAQ Nordic-börsen. De modeller som undersöktes inkluderade DeepLOB, DeepLOB-Attention, DeepLOB-seq2seq, DTNN och TCN. Det andra steget innefattade att köpa det svenska datasetet från Nasdaq Nordic, vilket tillhandahåller data om svenska aktier Limit Order Book (LOB). De två modellerna som visade bäst resultat i det första steget tränades sedan med detta dataset. Slutligen genomfördes en manipulation på de svenska orderböckerna för att undersöka hur dessa modeller påverkas. Resultatet utgjorde en tydlig bedömning av modellernas robusthet och pålitlighet när det gäller att förutse marknadsrörelser genom en omfattande jämförelse och analys av samtliga tester och deras resultat. Arbetet belyser även hur modellernas resultat påverkas av manipulativa handlingar. Dessutom framgår det hur valet av normaliseringsmetod påverkar modellernas resultat. / Over the years, intensive research has been conducted to enhance the capability of machine learning models to predict market movements. Despite this, during financial history, several events, such as the "Flash-crash," have impacted the market and had dramatic consequences for price movements. Therefore, it is crucial to examine how the models are affected by manipulative actions in the financial market to ensure their robustness and reliability in such situations. To carry out this work, the process has been divided into three steps. Firstly, a review of previous studies was conducted to identify the most robust models in the field. This was achieved by training the models on the FI-2010 dataset, which is a publicly available dataset for high-frequency trading of stocks on the NASDAQ Nordic stock exchange. The examined models included DeepLOB, DeepLOB-Attention, DeepLOB-seq2seq, DTNN, and TCN. The second step involved acquiring the Swedish dataset from Nasdaq Nordic, providing data on Swedish stock Limit Order Books (LOB). The two models that demonstrated the best results in the first step were then trained with this dataset. Finally, a manipulation was performed on the Swedish order books to investigate how these models would be affected. The result constituted a clear assessment of the models' robustness and reliability in predicting market movements through a comprehensive comparison and analysis of all tests and their results. The work also highlights how the models' outcomes are affected by manipulative actions. Furthermore, it becomes evident how the choice of normalization method affects the models' results.

Polyphonic or multiple music instrument detection is a difficult problem compared to detecting single or solo instruments in an audio recording. As music is time series data it be can modelled using sequence learning methods within deep learning. Recently, temporal convolutional networks (TCN) have shown to outperform conventional recurrent neural networks (RNN) on various sequence modelling tasks. Though there have been significant improvements in deep learning methods, data scarcity becomes a problem in training large scale models. Weakly labelled data is an alternative where a clip is annotated for presence or absence of instruments without specifying the times at which an instrument is sounding. This study investigates how TCN model compares to a Long Short-Term Memory (LSTM) model while trained on weakly labelled dataset. The results showed successful training of both models along with generalisation on a separate dataset. The comparison showed that TCN performed better than LSTM, but only marginally. Therefore, from the experiments carried out it could not be explicitly concluded if TCN is convincingly a better choice over LSTM in the context of instrument detection, but definitely a strong alternative. / Polyfonisk eller multipel musikinstrumentdetektering är ett svårt problem jämfört med att detektera enstaka eller soloinstrument i en ljudinspelning. Eftersom musik är tidsseriedata kan den modelleras med hjälp av sekvensinlärningsmetoder inom djup inlärning. Nyligen har ’Temporal Convolutional Network’ (TCN) visat sig överträffa konventionella ’Recurrent Neural Network’ (RNN) på flertalet sekvensmodelleringsuppgifter. Även om det har skett betydande förbättringar i metoder för djup inlärning, blir dataknapphet ett problem vid utbildning av storskaliga modeller. Svagt märkta data är ett alternativ där ett klipp kommenteras för närvaro av frånvaro av instrument utan att ange de tidpunkter då ett instrument låter. Denna studie undersöker hur TCN-modellen jämförs med en ’Long Short-Term Memory’ (LSTM) -modell medan den tränas i svagt märkta datasätt. Resultaten visade framgångsrik utbildning av båda modellerna tillsammans med generalisering i en separat datasats. Jämförelsen visade att TCN presterade bättre än LSTM, men endast marginellt. Därför kan man från de genomförda experimenten inte uttryckligen dra slutsatsen om TCN övertygande är ett bättre val jämfört med LSTM i samband med instrumentdetektering, men definitivt ett starkt alternativ.

No description available.

This dissertation focuses on the active tectonics of Afghanistan|slip-rate estimation of the Chaman fault and assessing seismic hazard in the Kabul basin. Afghanistan is a tectonically complex zone developed as a result of the collision between the Eurasian plate and the Indian plate to the southeast and the Arabian plate to the south. For seismic hazard mitigation, there is no large-scale active fault map in Afghanistan. I, therefore, mapped active and presumed active faults mainly based on interpretation of 1-arcsecond SRTM anaglyph images, and calculate the slip rate of the Chaman fautl based on 10Be TCN dating. / 博士(理学) / Doctor of Philosophy in Science / 同志社大学 / Doshisha University

No description available.