This report discusses some of the uses of algorithms within number theory. Topics examined include the applications of algorithms in the study of cryptology, the Euclidean Algorithm, prime generating functions, and the connections between algorithmic number theory and high school algebra. / text

Manufacturing today is no longer a single point of production activity but a chain of activities from the acquisition of raw materials to the delivery of products to customers. This chain is called supply chain. In this chain of activities, a generic pattern is: processing of goods (by manufacturers) and delivery of goods (to customers). This thesis concerns the scheduling operation for this generic supply chain. Two performance measures considered for evaluation of a particular schedule are: time and cost. Time refers to a span of the time that the manufacturer receives the request of goods from the customer to the time that the delivery tool (e.g. vehicle) is back to the manufacturer. Cost refers to the delivery cost only (as the production cost is considered as fi xed). A good schedule is thus with short time and low cost; yet the two may be in conflict. This thesis studies the algorithm for the supply chain scheduling problem to achieve a balanced short time and low cost. Three situations of the supply chain scheduling problem are considered in this thesis: (1) a single machine and multiple customers, (2) multiple machines and a single customer and (3) multiple machines and multiple customers. For each situation, di fferent vehicles characteristics and delivery patterns are considered. Properties of each problem are explored and algorithms are developed, analysed and tested (via simulation). Further, the robustness of the scheduling algorithms under uncertainty and the resilience of the scheduling algorithms under disruptions are also studied. At last a case study, about medical resources supply in an emergency situation, is conducted to illustrate how the developed algorithms can be applied to solve the practical problem. There are both technical merits and broader impacts with this thesis study. First, the problems studied are all new problems with the particular new attributes such as on-line, multiple-customers and multiple-machines, individual customer oriented, and limited capacity of delivery tools. Second, the notion of robustness and resilience to evaluate a scheduling algorithm are to the best of the author's knowledge new and may be open to a new avenue for the evaluation of any scheduling algorithm. In the domain of manufacturing and service provision in general, this thesis has provided an e ffective and effi cient tool for managing the operation of production and delivery in a situation where the demand is released without any prior knowledge (i.e., on-line demand). This situation appears in many manufacturing and service applications.

Soil moisture and temperature are two important soil parameters that influence many vital agronomic, environmental, engineering processes within the soil. Due to the difficulties arising when measuring these parameters in the field as well as the cost of instrumentation, many models that yield accurate and timely estimation of these parameters on a large scale have been developed as reliable and efficient alternatives. The Versatile Soil Moisture Budget model can be used to stimulate the vertical, one dimensional, water balance in a soil profile. Originally the model was designed to use air temperature and precipitation data to simulate soil water content within the root zone of a cereal crop. It has since undergone modifications and the model can now output, potential evapo-transpiration, actual evapo-transpiration, and surface temperature. The temperature algorithm simulates temperature at the soil surface and has not been rigorously tested for cropping systems. In this study, a simple empirical equation that simulates soil temperature at depth of up to 90 cm was introduced into the model. The model was evaluated and the accuracy of predicted soil moisture and temperature under both perennial and annual cropping systems were tested using two years of data collected at the University of Manitoba Research Station at Carman using soil water and temperature probes. The model’s accuracy in simulating soil moisture was also tested. Observed R2 comparing modelled temperature with observed was greater than 0.90 at the soil surface but decreased to about 0.40 at soil depth greater than 30-45 cm layer. The model was shown to be better at estimating soil temperature than soil moisture. The accuracy of the model was also shown to decrease with depth. These results can be used to improve soil temperature modeling at depth as well as improve farm management planning, irrigation schedules, nutrient management, fertilizer application and drought monitoring. / October 2015

The concept driver distraction is diffuse and no clear definition exists, which causes troubles when it comes to driver attention monitoring. This thesis takes an approach where eyetracking data from experienced drivers along with radar data has been used and analysed in an attempt to set up adaptive rules of how and how often the driver needs to attend to different objects in its surroundings, which circumvents the issue of not having a clear definition of driver distraction. In order to do this, a target tracking algorithm has been implemented that refines the output from the radar, subsequently used together with the eye-tracking data to in a statistical manner, in the long term, try to answer the question for how long is the driver allowed to look away in different driving scenarios? The thesis presents a proof of concept of this approach, and the results look promising.

We present a methodology for exploiting shared-memory parallelism within matrix computations by expressing linear algebra algorithms as directed acyclic graphs. Our solution involves a separation of concerns that completely hides the exploitation of parallelism from the code that implements the linear algebra algorithms. This approach to the problem is fundamentally different since we also address the issue of programmability instead of strictly focusing on parallelization. Using the separation of concerns, we present a framework for analyzing and developing scheduling algorithms and heuristics for this problem domain. As such, we develop a theory and practice of scheduling concepts for matrix computations in this dissertation. / text

In quiescent yeast, the widespread reorganization of cytosolic proteins into punctate has been observed (Narayanaswamy et al. 2009). We seek to better understand and describe this reorganization, which we hypothesize to be a protein aggregation phenomenon. To test this hypothesis, we examined mutant proteins (Ade4p protein variants) in yeast with predicted non-native aggregation propensities and measured their punctate formation kinetics. Monitoring punctate formation kinetics involved the validation of an automated quantification technique using an Amnis ImageStream imaging flow cytometer. The automated punctate counts were strongly correlated with the manual punctate counts, with usual R² values of 0.99 or better, but evaluated 50-fold more cells per run. Fitness evaluations of the mutant yeast in the form of growth curves and batch competition experiments revealed the slowed growth of the Ade4-1286 strain and the functional inequality to the wild type strain of the Ade4-mtoin2034, Ade4-mtoin2105, and Ade4-2800 strains in competition experiments, especially when the mutants were forced to generate their own adenine. Subsequent structural analysis of the mutant proteins revealed destabilizing mutations for 4 of the 6 mutant proteins with 2 of the mutations classified as significantly destabilizing ([delta][delta]G >2 kcal/mol). We concluded that the reduction in protein fitness was likely due to the destabilizing effects of the mutations. Evaluation of the punctate formation kinetics revealed little difference between strains in the rate of punctate formation. Further examination revealed the wild type Ade4p and all of the mutants (with the exception of the Ade4-1286 mutant) were predicted to have similar aggregation propensities according to a secondary aggregation predicting algorithm (Zyggregator, Pawar et al. 2005). Additionally, solvent accessibility calculations estimate ~3-19% of the side chain surface area to be solvent accessible, which indicates proximity of mutations to the protein surface. However, mutating buried amino acids likely would have generated a greater disturbance (Matthews 1993, Tokuriki et al. 2007). We concluded that the mutations, although destabilizing, altered the aggregation propensity very little. Deletion of chaperone proteins (Hsp82p, Hsc82p, and Ssa1p) revealed no difference in the Ade4-GFP punctate formation kinetics, although a slight kinetic difference was detected in the chaperone (Hsp82p) knockout, Gln1-GFP strain and the wild type strain. While further workup is necessary in the chaperone knockout, Gln1-GFP work, the initial results are promising and suggest the involvement of protein folding machinery in punctate formation. / text

Η εργασία πραγματέυεται τη δημιουργία βοηθητικών εργαλείων για την χρήση δεδομένων από αρχεία xml και την επεξεργασία τους με αλγορίθμους σμήνους για την εξαγωγή συμπερασμάτων. / The current post graduate thesis addresses the issue of school timetabling problem through the usage of PSO algorithms. In order to achieve greater efficiency complementary tool have been designed and developed.

Στην παρούσα διπλωματική εργασία παρουσιάζεται αρχικά η θεωρία της ψηφιακής επεξεργασίας σημάτων, βάσει της οποίας δημιουργήθηκαν οι τέσσερις αλγόριθμοι αποθορυβοποίησης ηχητικών σημάτων που μελετήθηκαν. Συγκεκριμένα μελετήθηκε η γραμμική φασματική αφαίρεση, ο αλγόριθμος Boll, ο αλγόριθμος Berouti και ο αλγόριθμος του Τσουκαλά, στον οποίο έγιναν και κάποιες τροποποιήσεις για βελτιστοποίηση της ποιότητας του αποθορυβοποιημένου σήματος. Πραγματοποιήθηκαν μετρήσεις SNR, NMR για τέσσερις διαφορετικές τιμές SNR ενθόρυβου σήματος: -5 , 0 , 5 και 10 dB. Το ενθόρυβο σήμα δημιουργήθηκε τεχνητά από ένα καθαρό σήμα ομιλίας σε ανηχοϊκό περιβάλλον με την χρήση λευκού θορύβου. Μέσω των μετρήσεων αυτών έγινε μια σύγκριση της απόδοσης των αλγορίθμων. / This thesis presents initially the theory of digital signal processing, in the bases of which the four noise reduction algorithms that are analysed, were created. In particular the algorithms that are analysed are the linear spectral subtraction, the Boll algorithm, the Berouti algorithm and the algorithm of Dionisios Tsoukalas, to which some optimizations were made for better quality of the noise reduced signal. SNR and NMR measurements were taken for four different noisy signal SNR values: -5, 0, 5 and 10 dB. The noisy signal was artificially created from a clean speech signal in an anechoic environment with the use of white noise. The performance of the algorithms has been compared through this measurements.

Η παρούσα μελέτη πραγματοποιήθηκε στα πλαίσια της διπλωματικής εργασίας «Ψηφιακός Επεξεργαστής και Γραφικά Υπολογιστών». Κατά τη διάρκεια αυτής της μελέτης είχαμε την ευκαιρία να προσεγγίσουμε και να κατανοήσουμε την διαδικασία που επιτελείται για την αναπαράσταση γραφικών στην οθόνη του υπολογιστή μας, του κινητού μας τηλεφώνου και κατ’ επέκταση να προτείνουμε νέες ιδέες και τεχνικές οι οποίες αποσκοπούν στην βελτίωση και αύξηση της αποδοτικότητας. Για το λόγο αυτό μελετώντας τους υπάρχοντες αλγόριθμους, καταλήξαμε στην πρόταση ενός νέου ο οποίος προορίζεται κυρίως για συσκευές απεικόνισης μικρού μεγέθους όπως τα κινητά τηλέφωνα και έχει ως βασικό σκοπό τη μείωση σύνθετων υπολογιστικών πράξεων και εξοικονόμηση ενέργειας. Στη παρούσα εργασία παρουσιάζεται η λειτουργία των υπαρχόντων αλγόριθμων καθώς και του προτεινόμενου. Πραγματοποιούνται συγκρίσεις μεταξύ αυτών προκειμένου να δούμε το ποσοστό βελτίωσης που υπεισέρχεται, ενώ αναπτύσσουμε και το υπόβαθρο για την υλοποίηση και εφαρμογή του προτεινόμενου αλγόριθμου στην πράξη. / -

Το αντικείμενο της διπλωματικής εργασίας είναι η εξαγωγή σχεδίου φλεβών από φωτογραφίες ανθρώπινων δακτύλων με τη χρήση ενός αλγορίθμου κορυφογραμμής. Στην αρχή, αναλύουμε κάποιες μεθόδους προσωπικού προσδιορισμού. Στη συνέχεια, αναφερόμαστε στη μέθοδο Hitachi και στις αρχές που την διέπουν. Κατόπιν, εξάγουμε τον αλγόριθμο της κορυφογραμμής και τον εφαρμόζουμε σε δεδομένες φωτογραφίες. Παραθέτουμε τα συμπεράσματα και τις παρατηρήσεις μας. Τέλος, στο παράρτημα εκθέτουμε το σύνολο του αλγορίθμου γραμμένο σε κώδικα MATLAB. / This study focus on the extraction of vein pattern from human fingers images with the use of an algorithm ridge. Firstly, we analyze certain methods of personal definition. Secondly, we make a reference to the method Hitachi and the principles is governed by. Then, we extract the algorithm ridge and we apply it on certain images. We cite our conclusions and comments. Finally, in the Appendix we present an overview of the algorithm using the MATLAB code.