Rob J Hyndman

Rob J Hyndmana, Heather Boothb and Farah Yas­meena aDepart­ment of Eco­no­met­rics & Busi­ness Stat­ist­ics, Mon­ash Uni­ver­sity, Clayton, Vic­toria, Aus­tralia. bThe Aus­tralian Demo­graphic & Social Research Insti­tute, Aus­tralian National Uni­ver­sity, Can­berra, ACT, Aus­tralia. Demo­graphy, to appear. Revised ver­sion: 20 April 2012. Abstract: When inde­pend­ence is assumed, fore­casts of mor­tal­ity for sub­pop­u­la­tions are almost always diver­gent in the long term. We pro­pose a method for coher­ent fore­cast­ing of mor­tal­ity rates for two or more sub­pop­u­la­tions, based on func­tional prin­cipal com­pon­ents mod­els of simple and inter­pretable func­tions of rates. The product-​​​​ratio func­tional fore­cast­ing method mod­els and fore­casts the geo­met­ric mean of sub­pop­u­la­tion rates and the ratio

(More)…

Shu Fan and Rob J Hyndman Revised 10 Janu­ary 2011 IEEE Trans­ac­tions on Power Sys­tems (2012), 27(1), 134–141. Abstract Short-​​​​term load fore­cast­ing is an essen­tial instru­ment in power sys­tem plan­ning, oper­a­tion and con­trol. Many oper­at­ing decisions are based on load fore­casts, such as dis­patch schedul­ing of gen­er­at­ing capa­city, reli­ab­il­ity analysis, and main­ten­ance plan­ning for the gen­er­at­ors. Over­es­tim­a­tion of elec­tri­city demand will cause a con­ser­vat­ive operation, which leads to the start-​​​​up of too many units or excess­ive energy pur­chase, thereby sup­ply­ing an unne­ces­sary level of reserve. On the con­trary, under­es­tim­a­tion may res­ult in a risky oper­a­tion, with insuf­fi­cient pre­par­a­tion of spin­ning reserve, caus­ing the sys­tem to oper­ate in a vul­ner­able region

(More)…

Bircan Erbas, Shahid Ullah, Rob J Hyndman, Michelle Scollo, Michael Abramson

BMC Med­ical Research Meth­od­o­logy, 2012, to appear.

Chronic Obstruct­ive Pul­mon­ary Dis­ease (COPD) is cur­rently the fifth lead­ing cause of death in Aus­tralia, and there are marked dif­fer­ences in mor­tal­ity trends between men and women. In this study, we have sought to model and fore­cast age related changes in COPD mor­tal­ity over time for men and women sep­ar­ately over the period 2006–2025.

Alysha M De Liv­era, Rob J Hyndman and Ralph D Snyder Journal of the Amer­ican Stat­ist­ical Asso­ci­ation (2011) 106(496), 1513–1527. Abstract A new innov­a­tions state space mod­el­ing frame­work, incor­por­at­ing Box-​​​​Cox trans­form­a­tions, Four­ier series with time vary­ing coef­fi­cients and ARMA error cor­rec­tion, is intro­duced for fore­cast­ing com­plex sea­sonal time series that can­not be handled using exist­ing fore­cast­ing mod­els. Such com­plex time series include time series with mul­tiple sea­sonal peri­ods, high fre­quency sea­son­al­ity, non-​​​​integer sea­son­al­ity and dual-​​​​calendar effects. Our new mod­el­ling frame­work provides an altern­at­ive to exist­ing expo­nen­tial smooth­ing mod­els, and is shown to have many advant­ages. The meth­ods for ini­tial­iz­a­tion and estim­a­tion, includ­ing like­li­hood eval­u­ation,

(More)…

John L Pearce, Jason Beringer, Neville Nich­olls, Rob J Hyndman, Pet­teri Uotila, and Nigel J Tapper

Atmo­spheric Envir­on­ment (2011), 45(1), 128–136.

The influ­ence of synoptic-​​scale cir­cu­la­tions on air qual­ity is an area of increas­ing interest to air qual­ity man­age­ment in regards to future cli­mate change. This study presents an ana­lysis where the dom­in­ant syn­op­tic ‘types’ over the region of Mel­bourne, Aus­tralia are determ­ined and linked to regional air quality.

Han Lin Shang, Heather Booth and Rob J Hyndman

Demo­graphic Research, 25(5), 173–214.

Using the age– and sex-​​specific data of 14 developed coun­tries, we com­pare the point and inter­val fore­cast accur­acy and bias of ten prin­cipal com­pon­ent meth­ods for fore­cast­ing mor­tal­ity rates and life expectancy.

Dav­ide Carta, Laura Vil­lan­ova, Stefano Cost­a­curta, Aless­andro Patelli, Irene Poli, Simone Vezzu, Paolo Sco­pece, Fabio Lisi, Kate Smith-​​​​Miles, Rob J Hyndman, Anita J. Hill, and Paolo Fal­caro Ana­lyt­ical Chem­istry (2011), 83(16), 6373–6380. ABSTRACT: In industry as well as many areas of sci­entific research, data col­lec­ted often con­tain a num­ber of responses of interest for a chosen set of explor­at­ory vari­ables. Optim­iz­a­tion of such mul­tivari­able mul­tire­sponse sys­tems is a chal­lenge well suited to genetic algorithms as global optim­iz­a­tion tools. One such example is the optim­iz­a­tion of coat­ing sur­faces with the required abso­lute and rel­at­ive sens­it­iv­ity for detect­ing ana­lytes using devices such as sensor arrays. High-​​​​throughput

(More)…