Tudor Pirvu

Rock-paper-scissors is a hand game used all over the world as a choosing method but compared to other methods it involves a bit of skill and experienced players can take advantage of amateur player’s behaviors in longer session games .
Since it is a zero-sum game that admits a pure Nash equilibrium in which both players use truly random strategies it is impossible to statistically beat a truly random opponent. However humans opponents can never be truly random so the aim of this project was to exploit this weakness and use computerized methods in order to gain advantage.
We managed to achieve a 72.2% win-loss rate in our experiments and we compared and contrasted different methods.
A unique approach of this project was the use of the Leap Motion device to make the collecting of data more realistic.

In this paper we formulate and study the stochastic game model that arises from the block chain mining activity. Since the participants in this proof-of-work scheme want to maximize their own utility,they may have reasons to diverge from the prescribed protocol and adopt selfish strategies, especially when their computational power is large.
The miners build a tree conjointly which is made out of a long trunk and occasionally short branches that later become stale. They receive rewards whenever their block ends up in the main path - (in bitcoin this is achieved at 100 blocks).
The aim of this project is to find relative computational power thresholds for different strategies in the following two games: the immediate release game, a game of complete information where miners release the blocks straight away after mining but they have the option of choosing which block to start mining from, and the strategic release game, a game with incomplete information where all blocks get announced after mining but not necessarily released immediately for others to start mining from.

The aim of the project was to designing a robust user interface integrating using the Leap Motion device and collecting video stream for post-processing by a human evaluator in order to diagnose or treat patients with Apraxia, a motor disorder caused by damage to the brain in which the individual has difficulty with the motor planning to perform tasks or movements when asked.
The final version of the project was used as a starting point for machine learning projects that aim to diagnose and devise a treatment plan without the need of a human evaluator