Dissertation topics suggested by the teacher.
______________________________________________________________________
APPEARANCE
Theses must conform in style to the practice of scientific literature. Figures and tables must be numbered and endowed with captions. Quotations are of utmost importance. Any claim that has been made by someone else must be ascribed to its author. For any information or data that has not been collected by the author, the subject that made it available must be credited.
In the text, quotes can either appear as "Blah blah blah (Smith and Johnson, 2013)" or "Smith and Johnson (2013) maintained that blah blah blah." All information regarding these quotes must appear in the list of references that concludes the thesis.
The list of references includes books and papers, as well as contents available in web sites, blogs or personal communications. As an example of references, please consult the INFORMS style.
______________________________________________________________________
INDEX:
- MASTER THESES IN COMPUTER SCIENCE APPLIED TO ECONOMICS/MANAGEMENT (see below)
- MASTER THESES IN ECONOMICS/MANAGEMENT (see page in Italian)
- BACHELOR THESES IN ECONOMICS/MANAGEMENT (see page in Italian)
____________________________________________________________________
MASTER THESES IN COMPUTER SCIENCE APPLIED TO ECONOMICS/MANAGEMENT
The origin of value
Why are we willing to pay for certain goods or services, wheras others are ignored? Why the value of certain innovations is discovered after so many years?
One possible answer is suggested by Steve Jobs' presentation of the first smartphone in 2007. In order to explain what it was, he said that it could be used to make a call, to listen to music, and to navigate the Internet. In other words, he explained the value of a new object by connecting it to the values provided by other objects.
I have a model of an artificial economy where the success of innovation depends on their ability to bridge between the bundles of knowledge that are associated with existing goods. As it stands, this artificial economy does not sustain it self. The thesis consists of experimenting with this model by exploring alternative architectures and by endowing its decision-makers with alternative behavioural rules until a sustainable artificial economy is generated.
This model is written in Java. Proficiency with Java is required.
Distributed representation of an organization's knowledge
Generally speaking, companies must strike a balance between exploration of novel knowledge and exploitation of available knowledge. Computational models of the exploration/exploitation trade-off typically represent an organization's knowledge by means of a string of zeroes and ones, a circumstance that obscures individual contributions as well as the existence of collective knowledge of which individuals are unaware.
This thesis consists of selecting or developing a recurrent neural network in order to reproduce organizational knowledge as stored in a distributed, content-addressable memory. Subsequently, reproducing and refining available knowledge about companies responding to environmental stimuli with alternative combinations of exploration and exploitation, as well as highlighting the interplay between collective knowledge and individual contributions.
This thesis requires proficiency with programming languages. Previous experience with neural networks would be very much welcome.
A generalist agent-based model of productive plants
While there exist several commercial models for productive plants in specific industries (e.g., automotive), Pietro Terna and Gianpaolo Jesi have developed a generalist Python-based model whose development had to be stopped because of lack of funds. Nevertheless, this model is potentially interesting because it might help understanding which factors influence the organizational learning curve.
The organizational leaerning curve, also known as progress function or progress ratio, describes the decrease of production time with cumulative production. It is particularly prominent in industries that are dominated by assembly operations, such as shipbuilding and airframes. Albeit it is essential to forecast production costs, no hint is available on what it depends. Engineers simply take past coefficients and hope they they will still hold, yet sometimes they do not.
The thesis consist of taking up the existing software (Python 2.7), assess its architecture, design a suitable architecture and program it in current Python. Knowledge of the Python programming language is a pre-requisite, some interest or experience with software architecture would be very much welcome.
A world of increasing returns to scale
Economic theory assumes that returns generally decrease with the scale of economic activities, a hypothesis that yields enormous simplifications because it implies that economies are at equilibrium. However, there exist indications that increasing returns to scale are rather the norm (i.e., the more you produce, the more money you make and, consequently, the more you produce), which implies that capitalist economies are rather like bubbling water, or exploding bombs.
The thesis consists of using a stock-flows agent-based macroeconomic model (model 02) to explore the consequences of generalized increasing returns to scale. Since this model is still underdevelopment, this thesis implies participating in its construction. Experience with Python is required.
Analysis of organizational routines in team sports
Team sports are an ideal setting to investigate organizational routines. The THEME software (History of Theme) is starting to be applied to this aim at the University of Iceland. The content and prospects of possible theses are still under negotiation with the authors of THEME.
