We show that often, there can be discrepancies between weather datasets that should be aligned. In these instances, we are able to train a neural network to predict a variable of one dataset from the other.
We demonstrate a method for improving IRL performance by allowing the expert agent to be careful. Carefulness means that the agent’s movements are less stochastic, but incur a higher movement penalty.
My undergraduate master’s dissertation in which I explore the applications of elliptic curves to cryptography. The work covers cryptanalysis, a key-agreement protocol for symmetric key encryption and methods for asymmetric key encryption and includes implementations of some of the discussed algorithms.
We derive an analogue of Ramanujan sums in the ring of polynomials over a finite field.