BEGIN:VCALENDAR
PRODID:-//TERMINALFOUR//SITEMANAGER V7.3//EN
VERSION:2.0
BEGIN:VEVENT
DTSTART:20170406T121500
LOCATION:Braamfontein Campus West Room 112, 1st Floor,The Liberty Actuarial Auditorium, Mathematical Sciences Laboratory Building
DESCRIPTION:Paul Mokilane, statistician at the Council for Scientific and Industrial Research and PhD candidate at 91心頭利, will present this seminar. Electricity demand forecasting is crucial not only in the day-to-day running of power systems, but also in system planning. Long-term forecasts are useful in capital planning.
<br />However, any prediction comes with uncertainties. Uncertainties in predictions imply that forecasts should ideally be probabilistic.
<br />Poor predictions could have far reaching consequences because an overestimate of long-term electricity demands could result in substantial wasted investment in the construction of excess power facilities, while underestimating of demands could result in insufficient generation and unmet future demand.
<br />The long term hourly electricity demand was forecasted using Quantile Regression (QR). In QR, the hourly electricity demand at different quantiles of the demand distribution which effectively described the full demand distribution is forecasted.
<br />&nbsp;
X-ALT-DESC;FMTTYPE=text/html:<strong>Paul Mokilane, statistician at the Council for Scientific and Industrial Research and PhD candidate at 91心頭利, will present this seminar. </strong><p>Electricity demand forecasting is crucial not only in the day-to-day running of power systems, but also in system planning. Long-term forecasts are useful in capital planning.</p>
<br /><p>However, any prediction comes with uncertainties. Uncertainties in predictions imply that forecasts should ideally be probabilistic.</p>
<br /><p>Poor predictions could have far reaching consequences because an overestimate of long-term electricity demands could result in substantial wasted investment in the construction of excess power facilities, while underestimating of demands could result in insufficient generation and unmet future demand.</p>
<br /><p>The long term hourly electricity demand was forecasted using Quantile Regression (QR). In QR, the hourly electricity demand at different quantiles of the demand distribution which effectively described the full demand distribution is forecasted.</p>
<br /><p>&nbsp;</p>
SUMMARY:Density forecasting of long-term electricity demand in South Africa using quantile regression
END:VEVENT
END:VCALENDAR