6533b824fe1ef96bd1280adb

RESEARCH PRODUCT

On the analysis of a new Markov chain which has applications in AI and machine learning

B. John OommenAnis YazidiOle-christopher Granmo

subject

VDP::Mathematics and natural science: 400::Mathematics: 410::Applied mathematics: 413InterleavingMarkov chainComputer sciencebusiness.industryStochastic processMarkov processVDP::Technology: 500::Information and communication technology: 550Machine learningcomputer.software_genreRandom walksymbols.namesakeChain (algebraic topology)symbolssortArtificial intelligencebusinesscomputer

description

Accepted version of an article from the conference: 2011 24th Canadian Conference on Electrical and Computer Engineering. Published version available from IEEE: http://dx.doi.org/10.1109/CCECE.2011.6030727 In this paper, we consider the analysis of a fascinating Random Walk (RW) that contains interleaving random steps and random "jumps". The characterizing aspect of such a chain is that every step is paired with its counterpart random jump. RWs of this sort have applications in testing of entities, where the entity is never allowed to make more than a pre-specified number of consecutive failures. This paper contains the analysis of the chain, some fascinating limiting properties, and some initial simulation results. The reader will find more detailed results in [12].

yearjournalcountryeditionlanguage
2011-05-01
http://hdl.handle.net/11250/137920