Μουρτάς Σπυρίδων
| Προσωπικές Πληροφορίες | Θέμα ΔΔ: | Intelligent Online Optimization Algorithms for Portfolio Analysis and Management |
| Επιβλέπων Καθηγητής: | Vasilios N. Katsikis, Department of Economics, National and Kapodistrian University of Athens, Greece | |
| Μέλος επιτροπής-1: | Predrag S. Stanimirovic, Department of Computer Science, University of Nis Serbia | |
| Μέλος επιτροπής-2: | Charalambos Tsitouras, General Department, National and Kapodistrian University of Athens, Greece | |
| Ηλεκτρονικό Ταχυδρομείο: | ||
| Προσωπική Ιστοσελίδα: | ||
| Περίληψη Διδακτορικής Διατριβής | Ελληνικά | |
| Τα μοντέλα βελτιστοποίησης διαδραματίζουν σημαντικό ρόλο στις οικονομικές αποφάσεις. Το πεδίο και ο σκοπός μας είναι να διερευνήσουμε εάν τα προβλήματα χρηματοοικονομικής βελτιστοποίησης μπορούν να έχουν online, επομένως πιο ρεαλιστική, λύση. Δηλαδή, τα προβλήματα αυτά πρέπει να μεταβάλλονται χρονικά ή να μετατραπούν πρώτα σε χρονικά μεταβαλλόμενα και στη συνέχεια θα τα λύσουμε χρησιμοποιώντας σύγχρονες μεθόδους τεχνητής νοημοσύνης. Έτσι μπορούμε να αποφύγουμε τους περιορισμούς της στατικής προσέγγισης, που συνήθως χρησιμοποιούνται στη βιβλιογραφία. Έχει αποδειχθεί ότι τα Linear-Variational Inequality based Primal-Dual Neural Networks (LVI-PDNN) προσεγγίζουν μια θεωρητική λύση όταν εφαρμόζονται ταυτόχρονα σε χρονικά μεταβαλλόμενα γραμμικά και quadratic προβλήματα που υπόκεινται σε ισότητα, ανισότητα και όρια. Πρόκειται για μια καινοτόμο προσέγγιση που περιλαμβάνει αξιόλογες τεχνικές από νευρωνικά δίκτυα για την παροχή της online λύσης σε προβλήματα οικονομικής βελτιστοποίησης. | ||
| Αγγλικά | ||
| Optimization models play a significant role in financial decisions. Our scope and purpose is to investigate whether financial optimization problems may have an online, therefore more realistic, solution. That is to say those problems must be time-varying or converted first into time-varying and then we will solve them by using modern Artificial Intelligence methods. We can thus avoid the limitations of the static approach, which is usually used in the literature. It has been shown that the Linear-Variational Inequality based Primal-Dual Neural Networks (LVI-PDNN), approaches a theoretical solution when applied simultaneously to time-varying linear and quadratic problems subject to equality, inequality and boundary constraints. This is a novel approach that comprises robust techniques from neural networks to provide the online solution to financial optimization problems. | ||
| Σύντομο Βιογραφικό | WORK EXPERIENCE Laboratories of Linear Maths Department of Economics of National and Kapodistrian University of Athens 2019 Key responsibilities included: • Teaching Linear Algebra • Teaching programming with Matlab Laboratories of Computer Science and Data Analysis Department of Economics of National and Kapodistrian University of Athens 2018 - 2019 Key responsibilities included: • Teaching advanced Microsoft Excel • Teaching Data Analysis using Microsoft Excel External Collaborator of e-socials.gr 2015 - 2017 Key responsibilities included: • Creating Ads • Schedule posts on social media Conservatories Patraiko, Patraiko Riou, Neo Odio Riou, Europio, Anastasopoulou, Messatidos 2007 - 2015 Key responsibilities included: • Teaching music theory • Teaching classical guitar • Teaching electric guitar • Teaching electric bass
EDUCATION PhD candidate of the Department of Economics of National and Kapodistrian University of Athens, Greece with dissertation title "Time-Varying Problems in Finance via Linear-Variational Inequality based Primal-Dual Neural Networks (LVI-PDNN)" 2019 - Current
MSc in Applied Economics and Finance with direction in Mathematical Finance and Risk Analysis of the Department of Economics of National and Kapodistrian University of Athens, Greece 2017 - 2019 BSc in Mathematics with direction in Computer Science and Computational Mathematics of the Department of Mathematics of the University of Patras, Greece 2005 – 2016 CERTIFICATES OF COMPETENCY ECPE - Certificate of Proficiency in English 2018 Web Developer Certificate (National and Kapodistrian University of Athens) 2017 Diploma in Counterpoint 2013 Diploma in Harmony 2011 Technical Computer Networks Certificate (Key-CERT IT Specialist) 2008 CERTIFICATES OF ATTENDANCE AUEB’s 16th Summer School on Risk Finance and Stochastics 2019 Attica Bank Innovation Days 2019 AUEB’s 15th Summer School in Stochastic Finance 2018 TECHNICAL SKILLS Programming: Fortran, C/C++, Prolog, Java, SQL, CSS, Html, Php, AJAX, JQUERY, Javascript, Python, Matlab, VBA Machine Learning: Deep Learning for Time Series Forecasting, Optimization Problems, Online Solution of Time-Varying Problems Neural Networks: Matlab Neural Network Toolbox, Tensorflow, Tensorflowjs, GNN, ZNN Excellent use of Eviews, SPSS, SAS, Latex, MS Office, CMS, Adobe Dreamweaver, Adobe InDesign, Adobe Photoshop Excellent use with most computer systems | |
| Ερευνητικές εργασίες - Δημοσιεύσεις | ● V.N. Katsikis, S. D. Mourtas, "A heuristic process on the existence of positive bases with applications to minimum-cost portfolio insurance in C [a,b]", Applied Mathematics and Computation349 (2019) 221-244. ● V.N. Katsikis, S. D. Mourtas, "ORPIT - a Matlab Toolbox for Option Replication and Portfolio Insurance in Incomplete Markets", Computational Economics(2019) 1-11. | |