Eventi come l’ultima crisi finanziaria sono una delle principali cause di perdite inattese negli investimenti finanziari. Infatti, durante una crisi finanziaria, la volatilità dei rendimenti azionari aumenta a causa degli shock dei mercati, incrementando la probabilità di perdita. Per fronteggiare gli effetti della crisi gli investitori , sfruttando possibili informazioni provenienti dai mercati, dovrebbero impostare con anticipo le proprie strategie di investimento e gestirli in modo appropriato per limitare gli effetti degli shock. In base a tali considerazioni si è analizzato, con tale studio, la capacità di predizione da parte di alcune variabili finanziarie/economiche ( BSEYD e Term Yield Spread) dell’andamento dei principali mercati europei ( Germania, Francia e Spagna) e si è definita un appropriata strategia di investimento per i periodi di crisi, costruendo un modello di portafoglio neutrale agli shock. I periodi analizzati sono stati rispettivamente : 1994-2013 e 2003-2014. I risultati hanno dimostrato che le variabili analizzate presentano diverso potere di predizione dei mercati considerati e che la perfomance del portafoglio neutrale agli shock di mercato è migliore di quella del portafoglio market neutral nei periodi di crisi. / Events like the last financial crisis are one of the principal causes of unexpected loss in investments. During a financial crisis period the stock return volatility increases for effect of internal and external shocks and the probability of reaching the expected return become lower , increasing the probability of loss. To face crisis effects, international investors should consider possible signals and information present in the market about possible crashes or declining period to set in advance their investment strategies and to manage them properly to limit the shock effects. On the basis of these considerations, with this study, we analyzed the predictive power of 2 economic/ financial variables( BSEYD and Term spread yield) on principal European stock markets (France, Germany and Spain) and defined a proper investment strategy for financial crisis period, building a portfolio model that is neutral to the international market shocks. The study has been conducted in two periods: 1994-2013 ( prediction analysis) and 2003-2014 (shock neutral portfolio mode). Results show different predictive power on the of the variables on the different markets analyzed and a better performance of the shock neutral portfolio than the market neutral portfolio strategy in declining period of the market.

本論文研究如何由觀測的選擇權市場價格還原風險中立機率測度（等價平賭測度）。首先建構選擇權投資組合的套利模型，其中假設選擇權為單期，到期日時的狀態為離散點且個數有限，並且對應同一標的資產且不同履約價格。若市場不存在套利機會時，可使用拉格朗日乘數法則將選擇權套利模型導出拉格朗日乘子的可行性問題。將可行性問題作為限制式重新建構線性規劃模型以還原風險中立機率測度，並且利用此風險中立機率測度評價選擇權的公正價格。最後，我們以台指選擇權（TXO）為例，驗證此模型的評價能力。 / This thesis investigates how to recover the risk-neutral probability (equivalent martingale measure) from observed market prices of options. It starts with building an arbitrage model of options portfolio in which the options are assumed to be in one-period time, finite discrete-states, and corresponding to the same underlying asset with different strike prices. If there is no arbitrage opportunity in the market, we can use Lagrangian multiplier method to obtain a Lagrangian multiplier feasibility problem from the arbitrage model. We employ the feasibility problem as the constraints to construct a linear programming model to recover the risk-neutral probability, and utilize this risk-neutral probability to evaluate the fair price of options. Finally, we take TXO as an example to verify the pricing ability of this model.

本研究探討於連續時間下，跨國投資者於匯率可預測下之最適投資決策問題。我們假設隨著時間改變，利用可預測之資訊動態修正投資決策。首先我們假設匯率可經由利率過程預測，探討相對風險趨避(CRRA)之投資經理人於跨國投資時之避險需求。研究方法是結合Cox與Huang (1989)之平賭方法與Lioui與Poncet (2003)於跨國投資所建構之財務模型。本研究歸納學習效果會影響匯率期望報酬，利用利率資訊會修正匯率過程之風險市場價值。最適投資決策因此受到調整因子之影響。因此投資人必須依照過濾進來的財務訊息(利率對匯率的改變)動態的調控持有之投資部位。最後，理論結果顯示投資部位必須針對可預測性下匯率避險效果作調整。 / In this study, we explore the effects of uncertainty about the exchange rate predictability on international portfolio choice in a continuous time setting. Uncertainty regarding to the predictive relation affects the optimal portfolio choice through dynamic learning, and leads to a state-dependent relation between the optimal portfolio choice and the investment time horizon. First we investigate the hedge demands in international portfolio management for constant relative risk averse investors where the exchange rate can be predicted by the change of interest rate. Then our approach is implemented through the use of the martingale methodology developed by Cox and Huang (1989) as proposed in the work of Lioui and Poncet (2003). Since the learning processes influence the premium of exchange rate movements, the crucial changes lie in the difference of market price of risk of the interest rate movements to the updated exchange rates. The constructed optimal investment strategy is influenced by the adjusted factors. Hence the investors should dynamically rebalance their holding portfolio according to the filtering mechanism. Finally, the theoretical results show that the adjustment for the optimal weights are required to reflect the prediction effects in hedging the exchange rate risks.

En una tesi anterior (Puig, 2004) es varen estudiar els aspectes mecànics vinculats al desfibratge i als consums energètics associats al procés de desintegració per acabar quantificant les forces implicades en el procés d'individualització dels components de diferents paperots. Aquest estudi, ha permès conèixer les condicions de desintegració que condueixen a un mínim consum energètic, aspecte complementari al reciclatge i destintatge del paperot però molt important per aconseguir un procés realment sostenible.Degut als requeriments del mercat, hi ha un altre aspecte o vessant realment important com és l'eliminació de tinta. Aquest procés, malgrat la complexitat industrial, es pot reduir o simplificar en dues etapes: la desintegració i la flotació. Ambdues, per raons d'eficàcia i condicionades pels agents químics que s'utilitzen, es realitzen en medi fortament alcalí (entorn d'un pH a 12). Aquest fet, genera una forta contaminació (DQO elevada) i per tant, és un procés poc respectuós amb el medi ambient.Els factors que condicionen treballar en medi alcalí són principalment, el despreniment de la tinta del suport, clarament accelerat en medi bàsic i fonamental per un bon destintatge i la utilització de tensioactius aniònics. Aquests tensioactius són sensibles al pH i la seva eficàcia disminueix en medi neutre o quasi neutre.Atès el que s'ha exposat, s'ha plantejat aquesta tesi que porta per títol "Aspectes Físics i Químics del Destintatge de Paper Revista mitjançant Tensioactius Catiònics", com un estudi del procés de destintatge de paper revista recuperat parant especial atenció a les variables físico-químiques i mecàniques del procés i a la seva influència en el despreniment i fragmentació de la tinta, així com en el procés d'eliminació de la tinta per flotació per tal d'avaluar la viabilitat d'un procés de destintatge en medi neutre enfront del destintatge alcalí convencional. S'ha utilitzat, com a paperot a destintar, el paper revista ja que aquest tipus de paper representa un percentatge elevat del paper recuperat a la Unió Europea. A més, aquest tipus de paper és utilitzat habitualment en les plantes de destintatge industrials per millorar les propietats de les suspensions destintades de paper de diari.La utilització de tensioactius catiònics és deguda a que la seva activitat no es veu afectada pel pH de la suspensió. Així, durant la comparació dels processos de destintatge en medi neutre i en medi alcalí, les eficàcies dels processos no es veuran influenciades pel grau d'ionització del tensioactiu.Donada l'absència d'estudis sobre la influència de les variables del procés en les mesures de concentració efectiva de tinta residual (ERIC) i ja que s'ha emprat aquest paràmetre per avaluar el despreniment i fragmentació de la tinta durant l'etapa de desintegració. En primer lloc, s'ha plantejat un estudi fonamental de la influència que poden tenir les variables del procés (granulometria, concentració de tinta i tipus de suport) en aquest paràmetre. Pel que fa a l'etapa de flotació, s'estudiarà l'efectivitat d'aquesta etapa, posant especial atenció a la capacitat del tensioactiu per eliminar tota la tinta despresa en l'etapa de desintegració. A més, donada la poca informació bibliogràfica referida als mecanismes de flotació amb tensioactius catiònics i la influència que té l'adsorció d'aquests agents per part de la suspensió paperera en l'etapa de flotació, s'ha dut a terme un estudi de l'adsorció dels tensioactius catiònics per part dels components d'una suspensió paperera.Per tant, la tesi realitzada té com a objectius: Estudi fonamental de les mesures de ERIC (concentració efectiva de tinta residual) amb suports específics. Estudi de la desintegració en medi neutre i comparació amb la desintegració clàssica. Modelització del desfibratge de paperot a l'etapa de desintegració. Estudi de l'eliminació de tinta per flotació en medi neutre i comparació amb la flotació clàssica. Estudi del mecanisme d'adsorció de tensioactius en suspensions de cel·lulosa sintètica i de paperot. / A previous thesis (Puig, 2004) have studied the mechanical aspects related to defibering and to energetic consumptions related to the disintegration process to end up quantifying the forces involved in the process of individualization of the different components of recovered papers. This study has permitted to know the disintegration conditions that lead to a minimum energetic consumption which is a complementary aspect to the recycling and deinking of waste paper but very important to get a really sustainable process.Due to the requests of the market, there is another really important aspect in the wastepaper recycling process which is the elimination of ink. Ink elimination, can be summarized in two unit operations: Disintegration and flotation. Both unit operations are strongly conditioned by the chemical agents used. Usually disintegration is carried out in a highly alkaline medium (pH 12). This fact, cause a high contamination in process waters (high COD) and therefore, the process is not environmentally friendly. The factors that condition the high alkaline medium are mainly ink detachment from the cellulosic support which is clearly accelerated by the alkaline medium, and the utilization of anionic surfactants. These surfactants are sensitive to the pH and its efficiency diminishes in neutral or almost neutral medium. So it has been proposed this thesis entitled "Aspectes Físics i Químics del Destintatge de Paper Revista mitjançant Tensioactius Catiònics" (Physico-chemical aspects of deinking of OMG using cationic surfactants) as a study of the viability of neutral deinking versus alkaline deinking.OMG has been used since this grade represents a high percentage of paper recovered in the European Union. Besides, this grade is used in industrial deinking plants to improve deinked ONP properties.Cationic surfactants are used because its properties are not affected by the pH of the suspension. So during the comparison of deinking processes in neutral and alkaline medium, the process efficiency will not be influenced by the surfactant ionization degree.OMG deinking study begins with a modelization of disintegration followed by an analysis of ink detachment and fragmentation achieved in neutral conditions and comparing them with the results obtained in alkaline conditions.Due to the absence of studies on the influence of the variables of the process in the measurements of effective residual ink concentration (ERIC) and since this parameter has been used to evaluate the detachment and fragmentation of ink during disintegration itself, it has been done a fundamental study of the influence of process variables such as ink particles size, ink concentration and type of the cellulosic support on this parameter.Regarding flotation stage, the effectiveness of this process will be studied focusing on the capability of the surfactant to eliminate all the ink detached during disintegration. Since deinking process is strongly influenced by surfactant adsorption on the different components of wastepaper suspensions and that there is a little information related to flotation mechanisms when cationic surfactants are used it has been accomplished a study of the adsorption of cationic surfactants by the components of a wastepaper suspension (cellulosic support and ink) and the results have been compared with the adsorption of an anionic surfactant in the same conditions.Therefore, the thesis has the following objectives: Fundamental study of ERIC measurements with specific supports. Study of neutral disintegration in comparison of traditional alkaline disintegration. Modelization of paper defibering during disintegration process. Study of ink elimination by means of a flotation process in neutral medium in comparison with a traditional alkaline flotation process. Study of the adsorption mechanism of surfactants by a wastepaper suspension.

In this doctoral dissertation three proposed approaches were evaluated using two databases of different languages, one American-English and one Greek. The proposed approaches were compared to the state-of-the-art models in the phone duration modelling task. The SVR model outperformed all the other individual models evaluated in this dissertation. Their ability to outperform all the other models is mainly based on their advantage of coping in a better way with high-dimensionality feature spaces in respect to the other models used in phone duration modelling, which makes them appropriate even for the case when the amount of the training data would be small respectively to the number of the feature set used. The proposed fusion scheme, taking advantage of the observation that different prediction algorithms perform better in different conditions, when implemented with SVR (SVR-fusion), contributed to the improvement of the phone duration prediction accuracy over that of the best individual model (SVR). Furthermore the SVR-fusion model managed to reduce the outliers in respect to the best individual model (SVR). Moreover, the proposed two-stage scheme using individual phone duration models as feature constructors in the first stage and feature vector extension (FVE) in the second stage, implemented with SVR (SVR-FVE), improved the prediction accuracy over the best individual predictor (SVR), and the SVR-fusion scheme and moreover managed to reduce the outliers in respect to the other two proposed schemes (SVR and SVR-fusion). The SVR two-stage scheme confirms in this way their advantage over all the other algorithms of coping well with high-dimensionality feature sets. The improved accuracy of phone duration modelling contributes to a better control of the prosody, and thus quality of synthetic speech. Furthermore, the first proposed method (SVR) was also evaluated on the phone duration modelling task in emotional speech, outperforming all the state-of-the-art models in all the emotional categories. Finally, perceptual tests were performed evaluating the impact of the proposed phone duration models to synthetic speech. The perceptual test for both the databases confirmed the results of objective tests showing the improvement achieved by the proposed models in the naturalness of synthesized speech. / Η παρούσα διδακτορική διατριβή πραγματεύεται προβλήματα που αφορούν στο χώρο της τεχνολογίας ομιλίας, με στόχο την μοντελοποίηση προσωδίας με χρήση τεχνικών μηχανικής μάθησης στα πλαίσια ουδέτερης και συναισθηματικής συνθετικής ομιλίας. Μελετήθηκαν τρεις καινοτόμες μέθοδοι μοντελοποίησης προσωδίας, οι οποίες αξιολογήθηκαν με αντικειμενικά τεστ και με υποκειμενικά τεστ ποιότητας ομιλίας για την συνεισφορά τους στην βελτίωση της ποιότητα της συνθετικής ομιλίας: Η πρώτη τεχνική μοντελοποίησης διάρκειας φωνημάτων, βασίζεται στην μοντελοποίηση με χρήση Μηχανών Υποστήριξης Διανυσμάτων (Support Vector Regression – SVR). Η μέθοδος αυτή δεν έχει χρησιμοποιηθεί έως σήμερα στην πρόβλεψη διάρκειας φωνημάτων. Η μέθοδος αυτή συγκρίθηκε και ξεπέρασε σε απόδοση όλες τις μεθόδους της επικρατούσας τεχνολογίας (state-of-the-art) στη μοντελοποίηση της διάρκειας φωνημάτων. Η δεύτερη τεχνική, βασίζεται στην μοντελοποίηση διάρκειας φωνημάτων με συνδυαστικό μοντέλο πολλαπλών προβλέψεων. Συγκεκριμένα, οι προβλέψεις διάρκειας φωνημάτων από ένα σύνολο ανεξάρτητων μοντέλων πρόβλεψης διάρκειας φωνημάτων χρησιμοποιούνται ως είσοδος σε ένα μοντέλο μηχανικής μάθησης, το οποίο συνδυάζει τις εξόδους από τα ανεξάρτητα μοντέλα πρόβλεψης και επιτυγχάνει μοντελοποίηση της διάρκειας φωνημάτων με μεγαλύτερη ακρίβεια, μειώνοντας επιπλέον και τα μεγάλα σφάλματα (outliers), δηλαδή τα σφάλματα που βρίσκονται μακριά από το μέσο όρο των σφαλμάτων. Η τρίτη τεχνική, είναι μια μέθοδος μοντελοποίησης διάρκειας φωνημάτων δύο σταδίων με κατασκευή νέων χαρακτηριστικών και επέκταση του διανύσματος χαρακτηριστικών. Συγκεκριμένα, στο πρώτο στάδιο, ένα σύνολο ανεξάρτητων μοντέλων πρόβλεψης διάρκειας φωνημάτων που χρησιμοποιούνται ως παραγωγοί νέων χαρακτηριστικών εμπλουτίζουν το διάνυσμα χαρακτηριστικών. Στο δεύτερο στάδιο, το εμπλουτισμένο διάνυσμα χρησιμοποιείται για να εκπαιδευτεί ένα μοντέλο πρόβλεψης διάρκειας φωνημάτων το οποίο επιτυγχάνει υψηλότερη απόδοση σε σχέση με όλες τις προηγούμενες μεθόδους, και μειώνει τα μεγάλα σφάλματα. Επιπλέον εφαρμόστηκε η πρώτη μέθοδος σε συναισθηματική ομιλία. Το προτεινόμενο SVR μοντέλο επιτυγχάνει την υψηλότερη απόδοση συγκρινόμενο με όλα τα state-of-the-art μοντέλα. Τέλος, πραγματοποιήθηκαν υποκειμενικά τεστ ποιότητας ομιλίας ώστε να αξιολογηθεί η συνεισφορά των τριών προτεινόμενων μεθόδων στη βελτίωση της ποιότητας της συνθετικής ομιλίας. Τα τεστ αυτά επιβεβαίωσαν την αξία των προτεινόμενων μεθόδων και τη συνεισφορά τους στη βελτίωση της ποιότητας στην συνθετική ομιλία.

Μελετήθηκε ο μεταβολισμός της γλυκερόλης στη ζύμη Yarrowia lipolytica ACA–DC 50109 με έμφαση στη μετατροπή της σε λιπίδια και κιτρικό οξύ, μεταβολικά προϊόντα που παρουσιάζουν ιδιαίτερο ενδιαφέρον για τη βιοτεχνολογία. Σε καλλιέργειες που πραγματοποιήθηκαν σε βιοαντιδραστήρα διαλείποντος έργου, επί πολλαπλώς περιοριστικού μέσου, διαπιστώθηκε η ύπαρξη τριών διακριτών φάσεων αύξησης που χαρακτηρίζονται από ιδιαίτερα μορφολογικά και βιοχημικά χαρακτηριστικά: η φάση βιοσύνθεσης κυτταρικής μάζας (κατά την οποία συντέθηκαν 4–4,5 g/l βιομάζας), η ελαιογόνος φάση (κατά την οποία πραγματοποιήθηκε συσσώρευση λιπιδίων 20–22% wt/wt επί ξηρής βιομάζας, 90% wt/wt των οποίων ήταν ουδέτερα) και η φάση παραγωγής κιτρικού οξέος (κατά την οποία εκκρίθηκαν στο περιβάλλον της αύξησης 14–30 g/l κιτρικού οξέος). Κατά τη διάρκεια των ανωτέρω φάσεων η ζύμη διήλθε από διάφορα μορφολογικά στάδια: μικρού μήκους αληθή μυκήλια και ψευδομυκήλια που κυριάρχησαν των κυττάρων ζύμης κατά τη φάση βιοσύνθεσης κυτταρικής μάζας, ευμεγέθη κύτταρα κατά τη φάση της ελαιογένεσης και μικρού μεγέθους κύτταρα ζύμης κατά τη φάση παραγωγής κιτρικού οξέος. Η γλυκερόλη διαπερνά την κυτταροπλασματική μεμβράνη με διευκολυνόμενη διάχυση και καταβολίζεται μέσω των αντιδράσεων της κινάσης της γλυκερόλης – GK και της NAD+ εξαρτώμενης αφυδρογονάσης της 3–P–γλυκερόλης. Την υψηλή ενεργότητα της NAD+ εξαρτώμενης ισοκιτρικής αφυδρογονάσης (NAD+–ICDH) κατά τη διάρκεια της φάσης βιοσύνθεσης κυτταρικής μάζας διαδέχθηκε σημαντική πτώση της ενεργότητάς της, επάγοντας τη λιπογένεση. Απρόσμενη αποδόμηση των αποθεματικών (ουδέτερων) λιπιδίων και σημαντική βιοσύνθεση γλυκολιπιδίων, σφιγγολιπιδίων και φωσφολιπιδίων – Ρ παρατηρήθηκε κατά τη διάρκεια της φάσης παραγωγής κιτρικού οξέος, φάση κατά την οποία η ενεργότητα της GK είχε μειωθεί σημαντικά ενώ η ενεργότητα της NAD+–ICDH είχε σχεδόν μηδενιστεί. Το ελαϊκό οξύ ήταν το κυριότερο λιπαρό οξύ ενώ η φωσφατιδυλχολίνη – PC το κύριο Ρ. Σε συνεχές σύστημα καλλιέργειας επί θρεπτικού υλικού περιοριστικού σε άζωτο, βιοσυντέθηκαν περιορισμένες μόνο ποσότητες λιπιδίων (~10% wt/wt, επί της ξηρής βιομάζας), γεγονός που μπορεί αποδοθεί στο ότι δεν υπήρχε μια περιοχή του ειδικού ρυθμού αραίωσης (D, h–1) στην οποία τα ένζυμα – κλειδιά που εμπλέκονται στη λιπογένεση (όπως η ΑΤΡ:κιτρική λυάση – ATP:CL και το μηλικό ένζυμο – ME) να παρουσιάζουν συγχρόνως υψηλές ενεργότητες, ενώ η ενεργότητα της NAD+–ICDH μειώθηκε, όχι όμως σημαντικά, στους χαμηλούς D. Η ενεργότητα της ATP:CL χαρακτηρίστηκε από υψηλές τιμές (60–300 Units/mg DW) σε D 0,033 h–1 ενώ οι μέγιστες τιμές ενεργότητας του ME (650 Units/mg DW) εμφανίστηκαν σε D=0,104 h–1. Τα λιπίδια της ζύμης ήταν περισσότερο ακόρεστα σε ενδιάμεσες τιμές D. Σε όλους τους D η φωσφατιδυλαιθανολαμίνη – PE, η φωσφατιδυλινοσιτόλη – PI και η PC αντιπροσωπεύουν τις κυριότερες κλάσεις των Ρ. Όσον αφορά τη μορφολογία της ζύμης, βρέθηκε ότι σε D<0,055 h–1 επικρατούσαν αληθή μυκήλια και ψευδομυκήλια ενώ σε D 0,055 h–1 παρατηρήθηκαν μόνο κύτταρα ζύμης. Σε πειράματα που πραγματοποιήθηκαν επί θρεπτικού υλικού περιοριστικού σε άζωτο, σε D=0,026 h–1, σε διαφορετικές συγκεντρώσεις διαλυμένου οξυγόνου – DO παρατηρήθηκε αυξημένο ποσοστό του κλάσματος των Ρ επί των ολικών λιπιδίων στις ακραίες σε τιμές DO ( 70% και 7%). Ανεξάρτητα των τιμών DO η PC ήταν η κλάση με το μεγαλύτερο ποσοστό, ακολουθούμενη από την PI και PE. Ειδικότερα το ποσοστό της ΡΕ παρουσιάστηκε ιδιαίτερα αυξημένο σε ενδιάμεσες τιμές DO (20% και 30%). Σε DΟ 50% επικρατούσαν αληθή μυκήλια και ψευδομυκήλια ενώ σε DΟ 50% εμφανίστηκαν στην καλλιέργεια περισσότερα κύτταρα ζύμης. Σε πειράματα που πραγματοποιήθηκαν σε D=0,026 h–1 βρέθηκε ότι ο περιορισμός της αύξησης από ιχνοστοιχεία όπως το μαγνήσιο και το ασβέστιο τα οποία εμπλέκονται σε πολλαπλές κυτταρικές λειτουργίες, είχαν δυσμενή επίδραση στη φυσιολογία της ζύμης, ωστόσο η σύσταση των λιπιδίων σε λιπαρά οξέα δεν επηρεάστηκε από τη φύση του περιοριστικού για την αύξηση παράγοντα. Η παρούσα διδακτορική διατριβή φιλοδοξεί να συμβάλει στη μελέτη της φυσιολογίας των ελαιογόνων μικροοργανισμών και στη χρήση της γλυκερόλης ως υποστρώματος σε μελλοντικές βιοτεχνολογικές εφαρμογές. / In this thesis the metabolism of glycerol in Yarrowia lipolytica ACA–DC 50109, with emphasis on glycerol conversion into value–added biotechnological products, such as single cell oils and citric acid, was studied. The growth of Y. lipolytica was studied in bioreactor batch cultures in multiple limited medium and three distinct phases were identified during growth cycle. In each phase, yeast cells were characterized by specific morphological and biochemical features: biomass formation phase (in which 4–4.5 g/l of biomass were synthesized), lipogenic phase (in which 20–22% lipids wt/wt in dry weight were accumulated in biomass, containing 90% wt/wt neutral lipids) and citric acid production phase (in which 14–30 g/l of citric acid were secreted in the growth environment). Distinct cellular forms of Y. lipolytica were developed during the above phases: in biomass formation phase short true mycelia and pseudo–mycelia were predominant while a few yeast–like cells were observed, in lipogenic phase large obese cells were predominant and in citric acid production phase cells size was diminished. Glycerol passes into the microbial cell by facilitated diffusion. Y. lipolytica successfully converts glycerol via phosphorylation pathway, in which glycerol kinase (GK) and glycerol–3–P–dehydrogenase are implicated. Though high activity of NAD+ dependent isocitric dehydrogenase (NAD+–ICDH) was detected during biomass formation phase, this activity was significantly decreased afterwards inducing lipogenesis. Surprisingly, storage (neutral) lipid turnover and synthesis of glycolipids, sphingolipids and phospholipids – Ρ simultaneously occurred with citric acid production, and happened when GK activity was considerably reduced and NAD+–ICDH activity was minimised. Oleic acid was the major fatty acid in all lipid fractions and phosphatidylcholine – PC was the main Ρ. In continuous culture in nitrogen limited medium Y. lipolytica accumulated low quantities of lipids (~10% w/w, in dry weight), maybe due to the fact that there was not a region of specific dilution rate (D, h–1) in which the key–enzymes that are implicated in lipogenesis (i.e. ΑΤΡ:citrate lyase – ATP:CL and malic enzyme – ME) presented simultaneously high activity while NAD+–ICDH activity was insignificantly decreased in low D. ATP:CL presented high activity (60–300 Units/mg DW) in D 0,033 h–1 while ME presented maximum activity (650 Units/mg DW) in D=0,104 h–1. Lipids were more unsaturated in intermediate D values while phosphatidylethanolamine – PE, phosphatidylinositol – PI and PC are the main Ρ classes. As far as the morphology is concerned, in D<0,055 h–1 short true mycelia and pseudo–mycelia were predominant in culture medium while in D 0,055 h–1 only yeast cells were observed. In experiments performed in nitrogen limited medium in D=0,026 h–1 in different dissolved oxygen – DO concentrations, it was found that in extreme DO values ( 70% and 7%) the percentage of P was increased. Independently the DO concentration PC was the main class followed by PI and PE. The morphology of Y. lipolytica was influenced by the different concentration of DO and it was observed that in DΟ 50% short true mycelia and pseudo–mycelia were predominant in culture medium while in DΟ 50% more yeast cells were appeared. In experiments performed in D=0,026 h–1, it was found that the absence of micronutrients from the growth medium, i.e. magnesium and calcium that are implicated in multiple cellular functions, had severe effects in yeast physiology, while the fatty acid composition of cellular lipids was not affected by the nature of the growth limiting factor. The present thesis aspires to contribute in the study of oleaginous microorganisms’ physiology and in use of glycerol as substrate in future biotechnological applications.

本研究探討VIX 期貨價格所隱含的資訊對於S&P 500 指數波動度預測的解釋力。過去許多文獻主要運用線性預測模型探討歷史波動度、隱含波動度和風險中立偏態對於波動度預測的資訊內涵。然而過去研究顯示，波動度具有長期記憶與非線性的特性，因此本文主要研究非線性預測模型對於波動度預測的有效性。本篇論文特別著重在不同市場狀態下(高波動與低波動)的實現波動度及隱含波動度異質自我迴歸模型(HAR-RV-IV model)。因此，本研究以考慮馬可夫狀態轉化下的異質自我迴歸模型(MRS-HAR model)進行實證分析。 本研究主要目的有以下三點: (1) 以VIX期貨價格所隱含的資訊提升S&P 500波動度預測的準確性。(2) 結合風險中立偏態與VIX期貨的資訊內涵，進一步提升S&P 500 波動度預測的準確性。(3) 考慮狀態轉換後的波動度預測模型是否優於過去文獻的線性迴歸模型。 本研究實證結果發現: (1) 相對於過去的實現波動度及隱含波動度，VIX 期貨可以提供對於預測未來波動度的額外資訊。 (2) 與其他模型比較，加入風險中立偏態和VIX 期貨萃取出的隱含波動度之波動度預測模型，只顯著提高預測未來一天波動度的準確性。 (3) 考慮狀態轉換後的波動度預測模型優於線性迴歸模型。 / This paper explores whether the information implied from VIX futures prices has incremental explanatory power for future volatility in the S&P 500 index. Most of prior studies adopt linear forecasting models to investigate the usefulness of historical volatility, implied volatility and risk-neutral skewness for volatility forecasting. However, previous literatures find out the long-memory and nonlinear property in volatility. Therefore, this study focuses on the nonlinear forecasting models to examine the effectiveness for volatility forecasting. In particular, we concentrate on Heterogeneous Autoregressive model of Realized Volatility and Implied Volatility (HAR-RV-IV) under different market conditions (i.e., high and low volatility state). This study has three main goals: First, to investigate whether the information extracted from VIX futures prices could improve the accuracy for future volatility forecasting. Second, combining the information content of risk-neutral skewness and VIX futures to enhance the predictive power for future volatility forecasting. Last, to explore whether the nonlinear models are superior to the linear models. This study finds that VIX futures prices contain additional information for future volatility, relative to past realized volatilities and implied volatility. Out-of-sample analysis confirms that VIX futures improves significantly the accuracy for future volatility forecasting. However, the improvement in the accuracy of volatility forecasts is significant only at daily forecast horizon after incorporating the information of risk-neutral skewness and VIX futures prices into the volatility forecasting model. Last, the volatility forecasting models are superior after taking the regime-switching into account.

Bibliography / Thesis (MA (General Linguistics))--University of Stellenbosch, 2011.

Fundação Carlos Chagas Filho de Amparo a Pesquisa do Estado do Rio de Janeiro / Como eventos de fissão induzida por nêutrons não ocorrem nas regiões nãomultiplicativas de reatores nucleares, e.g., moderador, refletor, e meios estruturais, essas regiões não geram potência e a eficiência computacional dos cálculos globais de reatores nucleares pode portanto ser aumentada eliminando os cálculos numéricos explícitos no interior das regiões não-multiplicativas em torno do núcleo ativo. É discutida nesta dissertação a eficiência computacional de condições de contorno aproximadas tipo albedo na formulação de ordenadas discretas (SN) para problemas de autovalor a dois grupos de energia em geometria bidimensional cartesiana. Albedo, palavra de origem latina para alvura, foi originalmente definido como a fração da luz incidente que é refletida difusamente por uma superfície. Esta palavra latina permaneceu como o termo científico usual em astronomia e nesta dissertação este conceito é estendido para reflexão de nêutrons. Este albedo SN nãoconvencional substitui aproximadamente a região refletora em torno do núcleo ativo do reator, pois os termos de fuga transversal são desprezados no interior do refletor. Se o problema, em particular, não possui termos de fuga transversal, i.e., trata-se de um problema unidimensional, então as condições de contorno albedo, como propostas nesta dissertação, são exatas. Por eficiência computacional entende-se analisar a precisão dos resultados numéricos em comparação com o tempo de execução computacional de cada simulação de um dado problema-modelo. Resultados numéricos para dois problemas-modelo com de simetria são considerados para ilustrar esta análise de eficiência. / As neutron fission events do not take place in the non-multiplying regions of nuclear reactors, e.g., moderator, reflector, and structural core, these regions do not generate power and the computational efficiency of nuclear reactor global calculations can hence be improved by eliminating the explicit numerical calculations within the non-multiplying regions around the active domain. Discussed here is the computational efficiency of approximate discrete ordinates (SN) albedo boundary conditions for two-energy group eigenvalue problems in X,Y geometry. Albedo, the Latin word for whiteness, was originally defined as the fraction of incident light reflected diffusely by a surface. This Latin word has remained the usual scientific term in astronomy and in this dissertation this concept is extended for the reflection of neutrons. The non-standard SN albedo substitutes approximately the reflector region around the active domain, as we neglect the transverse leakage terms within the nonmultiplying reflector. Should the problem have no transverse leakage terms, i.e., onedimensional slab geometry, then the offered albedo boundary conditions are exact. By computational efficiency we mean analyzing the accuracy of the numerical results versus the CPU execution time of each run for a given model problem. Numerical results to two symmetric test problems are shown to illustrate this efficiency analysis.

Conventional 2-level inverters have been quite popular in industry for drives applications. It used pulse width modulation techniques to generate a voltage waveform with high quality. For achieving this, it had to switch at high frequencies and also the switching is between 0 and Vdc. Also additional LC filters are required before feeding to a motor. 3-phase IM is the work horse of the industry. Several speed control techniques have been established namely the V/f control technique and for high performance, vector control is adopted. An electric drive system comprises of a rectifier, inverter, a motor and a load. each module is a topic by itself. This thesis work discusses the novel inverter topologies to overcome the demerits of a conventional 2-level inverter or even the basic multilevel topologies, for an electric drive. The word ‘multilevel’ itself signifies that inverter can generate more than two levels. The idea was first originated by Nabae, Takahashi and Akagi to bring an additional voltage level so that the waveform becomes a quasi square wave. This additional voltage level brought additional benefits in terms of reduced dv/dt and requirement of low switching frequency. But this was not without any cost. The inverter structure is slightly more complicated than a 2-level and also required more devices. But the advantage it gave was superior enough to such an extent that the above topology (popularly known as NPC) has become quite popular in industry. This topology was later modified to equalize the semiconductor losses among switches by replacing the clamping diodes with controllable switches and such topologies are popularly known as Active NPCs (ANPCs) because of the replacement of diodes with active switches. 3-level flying capacitors were then introduced where the additional voltage level is provided using charged capacitors. But this capacitor voltage has to be maintained at its nominal value during the inverter operation. An additional floating capacitor, which is an electrolytic capacitor is needed for this. Increasing the number of electrolytic capacitors reduces the reliability of the inverter drive since they are the weakest link in any inverters and its count has to be kept to the minimum. By using a H-bridge cell in each of the three phases, three voltage levels can be easily obtained.This is commonly known as Cascaded H-bridge (CHB) multilevel inverter. The above three topologies have been discussed with respect to generation of three pole voltage levels and these topologies are quite suited also. A higher number of voltage levels will reduce the switching frequency even lesser and also the dv/dt. On increasing the number of levels further and further, finally the inverter need not do any PWM switching and just generating the levels is sufficient enough for a good quality waveform and also low dv/dt. But when the above topologies are scaled for more than three voltage levels, all of them suffer serious drawbacks which is briefly discussed below. The diode clamped inverter (known as NPC if it is 3-level), when extended to more than three levels suffers from the neutral point balancing issue and also the count of clamping diodes increase drastically. FC inverters, when extended beyond 3-level, the number of electrolytic capacitors increases and also balancing of these capacitors to their nominal voltages becomes complicated. In the case of multilevel CHB, when extended beyond 3-level, the requirement of isolated DC sources also increases. To generate isolated supplies, phase shifting transformer and 8, 12 or 24 pulse diode rectifier is needed which increases the weight , size and cost of the drive. Therefore its application is limited. In this thesis, the aim is to develop a novel method to develop a multilevel inverter without the drawbacks faced by the basic multilevel topologies when scaled for higher number of voltage levels. This is done through stacking the basic or hybrid combination of these basic multilevel topologies through selector switches. This method is experimentally verified by stacking two 5-level inverters through a 2-level selector switch (whose switching losses can be minimized through soft cycle commutation). This will generate nine levels.Generating 9-levels through scaling the basic topologies is disadvantageous, the comparison table is provided in the thesis. This is true for any higher voltage level generation. Each of the above 5-level inverter is developed through cascading an FC with a capacitor fed H-bridge. The device count can be reduced by making the FC-CHB module common to the selector switches by shifting the selector switches between the DC link and the common FC-CHB module. Doing so, reduces the modular feature of the drive but the device count can be reduced. The FFT plot at different frequencies of operation and the switching losses of the different modules-FC, CHB and the selector switches are also plotted for different frequencies of operation. The next step is to check whether this method can be extended to any number of stackings for generation of more voltage levels. For this, a 49-level inverter is developed in laboratory by stacking three 17-level inverters. Each of the 17-level inverter is developed by cascading an FC with three CHBs. When there are 49 levels in the pole voltage waveform, there is no need to do any regular PWM since the output waveform will be very close to a sine wave even without any PWM switching. The technique used is commonly known in literature as Nearest Level Control (NLC). This method of stacking and cascading has the advantage that the FC and the CHB modules now are of very low voltages and the switching losses can be reduced. The switching losses of the different modules are calculated and plotted for different operating frequencies in the thesis. To reduce the voltages of the modules further, a 6-phase machine has been reconfigured as a 3-phase machine, the advantage being that now the DC link voltage requirement is half of that needed earlier for the same power. This further reduces voltages of the modules by half and this allows the switches to be replaced with MOSFETs, improving the efficiency of the drive. This topology is also experimentally verified for both steady state and transient conditions. So far the research focussed on a 3-phase IM fed through a stacked MLI. It can be observed that a stacked MLI needs as many DC sources as the number of stackings. A 6-phase machine apart from reduced DC link voltage requirement, has other advantages of better fault tolerant capability and better space harmonics. They are serious contenders for applications like ship propulsion, locomotive traction, electric vehicles, more electric aircraft and other high power industrial applications. Using the unique property of a 6-phase machine that its opposite windings always draw equal and opposite current, the neutral point (NP) (formed as a result of stacking two MLIs) voltage can be balanced. It was observed that the net mid point current drawn from the mid point can be made zero in a switching interval. It was later observed that with minimal changes, the mid point current drawn from the NP can be made instantaneously zero and the NP voltage deviation is completely arrested and the topology needs only very low capacity series connected capacitors energized from a single DC link. This topology is also experimentally verified using the stacked 9-level inverter topology discussed above but now for 6-phase application and experimental results are provided in the thesis. Single DC link enables direct back to back conversion and power can be fed back to the mains at any desired power factor. All the experimental verification is done on a DSP (TMS320F28335) and FPGA (Spartan 3 XCS3200) platform. An IM is run using V/f control scheme and the above inverter topologies are used to drive the motor. The IGBTs used are SKM75GB123D for the stacked 9-level inverter in the 3-phase and 6-phase experiments. For the 49-level inverter experiment, MOSFETs-IRF260N were used. Both steady state and transient results ensure that the proposed inverter topologies are suitable for high power applications.