Modeling together with Forecasting Electricity Loads together with Prices: H5N1 Statistical Approach
ISBN: 978-0-470-05753-7
192 pages
Dec 2006
Book of the day: http://www.wiley.com/WileyCDA/WileyTitle/productCd-047005753X.html
=============================================
Source: https://branchesofmechanicalengineering.blogspot.com//search?q=whos-who-in-energy-forecasting-rafal-weron
Monday, July 31, 2017
Who's Who inward Energy Forecasting: Rafał Weron
Prof. Rafał Weron is the Head of the Economic Modeling Group inward the Department of Operations Research at Wroclaw University of Science together with Technology. He has thus many accomplishments inward unloose energy forecasting, such every bit his renowned mass on charge together with cost forecasting, many widely cited papers, together with of course of written report those talented students. Recently he won the IIF-Hong Award for his IJF review newspaper on electricity cost forecasting.
What brought y'all to the unloose energy field, especially charge together with cost forecasting?
I would similar to state that a good thought-out determination ... but realistically ... it was to a greater extent than luck together with coincidence. I was the correct individual at the correct time.
Towards the halt of my PhD studies, dorsum inward 1998, I started looking for a novel surface area of research. I studied mathematics, my PhD was from something stylish inward the 1990s – math finance. But it was likewise theoretical for me, likewise far from existent applications. Finance itself was better. Yet, doing top grade finance enquiry together with publishing inward top-tier finance journals was (nearly) impossible for someone working inward a one-time East Block province dorsum then. On the other hand, I liked the liberty of the Academia together with didn’t desire to piece of work every bit a quant.
Then inward the belatedly 1990s, ability marketplace deregulation started spreading throughout Europe together with the US. The Polish Power Exchange opened inward mid-2000, but at that topographic point were rattling few people who knew what ability trading was about. Mathematicians together with economists were fix dorsum past times the technical aspects of ability marketplace operations, engineers had no economical training. I saw this every bit an opportunity. With a audio grooming inward statistics/time serial analysis, one-half a yr spent every bit a trainee inward an investment depository fiscal establishment together with basic cognition almost ability organisation economic science gained during a projection run past times the Hugo Steinhaus Center for the Polish Power Grid inward 1995-1996, I had a caput start compared to swain colleagues. And I was eager to piece of work hard.
But my starting fourth dimension “energy” papers were non on forecasting. Rather on information analytics, adventure administration together with derivatives pricing, focusing to a greater extent than on the mid-term horizon. I became interested short-term forecasting a chip later, around 2003-2004, when it became apparent that the day-ahead marketplace was “the marketplace” for electricity trading, non the derivatives markets.
Tell us to a greater extent than almost your cost forecasting review paper. Why together with how did y'all write that 52-page article?
The invitation from Rob Hyndman to write the review newspaper came out of the blue, around mid-2013. Ask him why he approached me inward the starting fourth dimension place. Prior to that I had alone i newspaper inward IJF, inward a special number on Energy Forecasting published inward 2008.
But it was a welcome invitation. H5N1 few years create got passed since I published my 2006 Wiley mass on modeling together with forecasting electricity loads together with prices. And I create got been gathering material
for a revised version. So I agreed happily together with said that I would submit a draft past times the halt of the year. This turned out to live on impossible … well, I started working on it likewise late, sometime inward early on Dec 2013, after Rob had ship me a reminder electronic mail ;-)
I wanted my review to live on self-contained together with rather complete. I loathe these review/survey papers which only refer dozens of articles without genuinely analyzing together with thoroughly comparison the results using the same measures. So I wanted to include my ain empirical examples. This meant a lot of coding together with information analysis. No wonder it took me 2 total months to complete. But the consequence surprised fifty-fifty me – it was similar a modest mass – the draft had 88 pages inward the criterion elsarticle Latex page layout. I was certain Rob would tell me to cutting it inward one-half …
What's your proudest accomplishment inward forecasting?
A famous Polish mathematician, Hugo Steinhaus, used to state that his greatest regain was … Stefan Banach. Yes, the same Banach known inward mathematics for “Banach spaces”, i of the founders of modern functional analysis. I also similar to recall that my greatest accomplishments are my bright students. One of them – Jakub Nowotarski – graduated this June, amongst distinction. Jakub’s enquiry output was outstanding, non alone for a PhD student. The Quantile Regression Averaging method, which turned out to live on a top performer inward the GEFCom2014 competition, was 90% his idea. He would create got easily received a habilitation ( tenure) inside a yr or two, if alone he decided to remain inward the Academia. But I am working right away amongst 2 gifted BSc students – Grzegorz Marcjasz together with Bartosz Uniejewski. Someday I may live on able to state the same almost them.
Do y'all piece of work amongst companies to improve their forecasting practice?
I create got had dissimilar episodes inward my life, some to a greater extent than academic, some less. Over the final 2 decades I create got been periodically engaged every bit a consultant to financial, unloose energy together with software applied scientific discipline companies. And yes, I create got worked amongst utilities together with ability generators on improving their charge together with cost forecasting techniques together with adventure administration systems. The hype on unloose energy forecasting inward Poland was betwixt 2000 together with 2006. Then a serial of mergers changed the landscape – the iv large companies that remained were non that interested inward developing in-house solutions anymore. So my recent developments inward forecasting are to a greater extent than academic inward nature.
Is at that topographic point a key maiden or exciting projection y'all are working?
Together amongst Florian Ziel nosotros are working on a mass for CRC Press that volition supervene upon my 2006 Wiley book. The tentative championship is “Forecasting Electricity Prices: H5N1 Guide to Robust Modeling”. It is scheduled to live on out inward 2018.
The mass volition start amongst a chapter on the fine art of forecasting, introducing the basic notions of (energy) forecasting. We volition travel on amongst a chapter on the markets for electricity together with hash out the products traded there. Then the 3 primary chapters volition follow: “Forecasting for Beginners” – which volition innovate a few unproblematic models together with demonstrate how point, probabilistic together with path forecasts tin live on computed for them, “Evaluating Models together with Forecasts” – which is a rattling important, but withal underdeveloped surface area inward unloose energy forecasting, together with “Forecasting for Experts” – which volition hash out a number of to a greater extent than advanced concepts, similar regime-switching, shrinkage, characteristic selection, non-linear together with key models.
What's your forecast for the adjacent 10 years of unloose energy forecasting field?
This is a tough question. When writing my 2014 IJF review I came upward amongst 5 directions inward which electricity cost forecasting would or should evolve over the adjacent decade: (1) a amend handling of seasonality together with exercise of fundamentals, (2) going beyond signal forecasts, (3) to a greater extent than extensive exercise of forecast combinations, (4) evolution of multivariate models, together with (5) to a greater extent than thorough forecast evaluation. Out of these, I recall that the to the lowest degree has been done since thus inward the context of multivariate models. This is non a surprise, multivariate models are much to a greater extent than demanding, non alone conceptually but also computationally. But I create believe that they create got a lot to offer. Also Bayesian methods may encounter to a greater extent than extensive use, especially inward probabilistic forecasting.
Another administration that I recall may larn of import inward the nigh time to come is “path forecasting”. Currently, inward a vast bulk of charge or cost forecasting papers alone marginal (i.e., at i signal inward time) distributions are considered, either inward a signal or probabilistic context. But the forecast for hr ix should non live on independent from the i for hr 8. If nosotros predict a cost drib below a “normal” grade for hr 8 tomorrow, thus it is quite probable that the cost for hr ix volition also live on lower than nether “normal” circumstances. Our forecasting models should create got this into account.
What else create y'all create inward the academic basis other than unloose energy forecasting?
Currently it’s 75% percentage unloose energy forecasting together with 25% agent based modeling, but also related to unloose energy markets – diffusion of innovations, similar dynamic tariffs or pro-ecological behavior. In the non thus distant past times I create got worked on long-range dependence, adventure administration together with derivatives pricing, also exterior the Academia.
What's fun almost your job?
Everything. The sleepless nights spent on writing papers, the discussions amongst the reviewers together with editors that I am correct together with they are wrong, together with – most importantly – dozens of emails together with skype calls I telephone commutation each 24-hour interval amongst my students when working on a novel enquiry idea.
How create y'all pass your complimentary time?
Working. No, this would live on an exaggeration … but alone a modest one. H5N1 researcher is never on a vacation – best ideas don’t come upward during my business office hours, they tend to popular upward unexpectedly. But I crusade my best non to live on a 100% workaholic. I similar mountains – both hiking together with skiing. From sports – playing volleyball, badminton together with squeeze – non that I’m a expert player, but I similar it ;-)
What brought y'all to the unloose energy field, especially charge together with cost forecasting?
I would similar to state that a good thought-out determination ... but realistically ... it was to a greater extent than luck together with coincidence. I was the correct individual at the correct time.
Towards the halt of my PhD studies, dorsum inward 1998, I started looking for a novel surface area of research. I studied mathematics, my PhD was from something stylish inward the 1990s – math finance. But it was likewise theoretical for me, likewise far from existent applications. Finance itself was better. Yet, doing top grade finance enquiry together with publishing inward top-tier finance journals was (nearly) impossible for someone working inward a one-time East Block province dorsum then. On the other hand, I liked the liberty of the Academia together with didn’t desire to piece of work every bit a quant.
Then inward the belatedly 1990s, ability marketplace deregulation started spreading throughout Europe together with the US. The Polish Power Exchange opened inward mid-2000, but at that topographic point were rattling few people who knew what ability trading was about. Mathematicians together with economists were fix dorsum past times the technical aspects of ability marketplace operations, engineers had no economical training. I saw this every bit an opportunity. With a audio grooming inward statistics/time serial analysis, one-half a yr spent every bit a trainee inward an investment depository fiscal establishment together with basic cognition almost ability organisation economic science gained during a projection run past times the Hugo Steinhaus Center for the Polish Power Grid inward 1995-1996, I had a caput start compared to swain colleagues. And I was eager to piece of work hard.
But my starting fourth dimension “energy” papers were non on forecasting. Rather on information analytics, adventure administration together with derivatives pricing, focusing to a greater extent than on the mid-term horizon. I became interested short-term forecasting a chip later, around 2003-2004, when it became apparent that the day-ahead marketplace was “the marketplace” for electricity trading, non the derivatives markets.
Tell us to a greater extent than almost your cost forecasting review paper. Why together with how did y'all write that 52-page article?
The invitation from Rob Hyndman to write the review newspaper came out of the blue, around mid-2013. Ask him why he approached me inward the starting fourth dimension place. Prior to that I had alone i newspaper inward IJF, inward a special number on Energy Forecasting published inward 2008.
But it was a welcome invitation. H5N1 few years create got passed since I published my 2006 Wiley mass on modeling together with forecasting electricity loads together with prices. And I create got been gathering material
for a revised version. So I agreed happily together with said that I would submit a draft past times the halt of the year. This turned out to live on impossible … well, I started working on it likewise late, sometime inward early on Dec 2013, after Rob had ship me a reminder electronic mail ;-)
I wanted my review to live on self-contained together with rather complete. I loathe these review/survey papers which only refer dozens of articles without genuinely analyzing together with thoroughly comparison the results using the same measures. So I wanted to include my ain empirical examples. This meant a lot of coding together with information analysis. No wonder it took me 2 total months to complete. But the consequence surprised fifty-fifty me – it was similar a modest mass – the draft had 88 pages inward the criterion elsarticle Latex page layout. I was certain Rob would tell me to cutting it inward one-half …
What's your proudest accomplishment inward forecasting?
A famous Polish mathematician, Hugo Steinhaus, used to state that his greatest regain was … Stefan Banach. Yes, the same Banach known inward mathematics for “Banach spaces”, i of the founders of modern functional analysis. I also similar to recall that my greatest accomplishments are my bright students. One of them – Jakub Nowotarski – graduated this June, amongst distinction. Jakub’s enquiry output was outstanding, non alone for a PhD student. The Quantile Regression Averaging method, which turned out to live on a top performer inward the GEFCom2014 competition, was 90% his idea. He would create got easily received a habilitation ( tenure) inside a yr or two, if alone he decided to remain inward the Academia. But I am working right away amongst 2 gifted BSc students – Grzegorz Marcjasz together with Bartosz Uniejewski. Someday I may live on able to state the same almost them.
Do y'all piece of work amongst companies to improve their forecasting practice?
I create got had dissimilar episodes inward my life, some to a greater extent than academic, some less. Over the final 2 decades I create got been periodically engaged every bit a consultant to financial, unloose energy together with software applied scientific discipline companies. And yes, I create got worked amongst utilities together with ability generators on improving their charge together with cost forecasting techniques together with adventure administration systems. The hype on unloose energy forecasting inward Poland was betwixt 2000 together with 2006. Then a serial of mergers changed the landscape – the iv large companies that remained were non that interested inward developing in-house solutions anymore. So my recent developments inward forecasting are to a greater extent than academic inward nature.
Is at that topographic point a key maiden or exciting projection y'all are working?
Together amongst Florian Ziel nosotros are working on a mass for CRC Press that volition supervene upon my 2006 Wiley book. The tentative championship is “Forecasting Electricity Prices: H5N1 Guide to Robust Modeling”. It is scheduled to live on out inward 2018.
The mass volition start amongst a chapter on the fine art of forecasting, introducing the basic notions of (energy) forecasting. We volition travel on amongst a chapter on the markets for electricity together with hash out the products traded there. Then the 3 primary chapters volition follow: “Forecasting for Beginners” – which volition innovate a few unproblematic models together with demonstrate how point, probabilistic together with path forecasts tin live on computed for them, “Evaluating Models together with Forecasts” – which is a rattling important, but withal underdeveloped surface area inward unloose energy forecasting, together with “Forecasting for Experts” – which volition hash out a number of to a greater extent than advanced concepts, similar regime-switching, shrinkage, characteristic selection, non-linear together with key models.
What's your forecast for the adjacent 10 years of unloose energy forecasting field?
This is a tough question. When writing my 2014 IJF review I came upward amongst 5 directions inward which electricity cost forecasting would or should evolve over the adjacent decade: (1) a amend handling of seasonality together with exercise of fundamentals, (2) going beyond signal forecasts, (3) to a greater extent than extensive exercise of forecast combinations, (4) evolution of multivariate models, together with (5) to a greater extent than thorough forecast evaluation. Out of these, I recall that the to the lowest degree has been done since thus inward the context of multivariate models. This is non a surprise, multivariate models are much to a greater extent than demanding, non alone conceptually but also computationally. But I create believe that they create got a lot to offer. Also Bayesian methods may encounter to a greater extent than extensive use, especially inward probabilistic forecasting.
Another administration that I recall may larn of import inward the nigh time to come is “path forecasting”. Currently, inward a vast bulk of charge or cost forecasting papers alone marginal (i.e., at i signal inward time) distributions are considered, either inward a signal or probabilistic context. But the forecast for hr ix should non live on independent from the i for hr 8. If nosotros predict a cost drib below a “normal” grade for hr 8 tomorrow, thus it is quite probable that the cost for hr ix volition also live on lower than nether “normal” circumstances. Our forecasting models should create got this into account.
What else create y'all create inward the academic basis other than unloose energy forecasting?
Currently it’s 75% percentage unloose energy forecasting together with 25% agent based modeling, but also related to unloose energy markets – diffusion of innovations, similar dynamic tariffs or pro-ecological behavior. In the non thus distant past times I create got worked on long-range dependence, adventure administration together with derivatives pricing, also exterior the Academia.
What's fun almost your job?
Everything. The sleepless nights spent on writing papers, the discussions amongst the reviewers together with editors that I am correct together with they are wrong, together with – most importantly – dozens of emails together with skype calls I telephone commutation each 24-hour interval amongst my students when working on a novel enquiry idea.
How create y'all pass your complimentary time?
Working. No, this would live on an exaggeration … but alone a modest one. H5N1 researcher is never on a vacation – best ideas don’t come upward during my business office hours, they tend to popular upward unexpectedly. But I crusade my best non to live on a 100% workaholic. I similar mountains – both hiking together with skiing. From sports – playing volleyball, badminton together with squeeze – non that I’m a expert player, but I similar it ;-)