Honours

I did hon­ours in com­puter sci­ence in 2007, under the super­vi­sion of Dr Tasos Viglas at the School of Inform­a­tion Tech­no­lo­gies at the Uni­ver­sity of Sydney.

Descrip­tion

In a net­work, greedy, inde­pend­ent agents aim to min­im­ise their own per­sonal cost (such as travel time between source and des­tin­a­tion) without regard to wider, soci­etal impacts of their beha­viour. The inef­fi­ciency due to this beha­viour can be stud­ied through such meas­ures as the price of anarchy, which is the ratio of the cost of the worst-​​case Nash equi­lib­rium to that of the optimal flow. One applic­a­tion of this game the­or­etic ana­lysis is in allow­ing net­work oper­at­ors to charge users equit­ably and prof­it­ably for mul­tic­ast traffic sent through their net­works, because mul­tic­ast traffic along a link can­not be simply attrib­uted to one par­tic­u­lar user. The aim of the pro­ject is to extend and modify exist­ing mod­els of mul­tic­ast pri­cing, to improve the res­ult­ant soci­etal cost even with greedy, inde­pend­ent agents and their applic­ab­il­ity to real-​​world mul­tic­ast uses. The­or­et­ical prop­er­ties, such as time and net­work over­head com­plex­ity, will be examined to determ­ine the tract­ab­il­ity of the models.

Doc­u­ments

For INFO4990 (IT Research Meth­ods) course

Other