Thursday, June 20, 2019

Energy Forecasting in the Big Data World

All papers for the International Journal of Forecasting special section on energy forecasting in the big data world have been published online. Out of 14 papers collected for this special section, eight are from GEFCom2017 documenting winning methods, while the other six non-GEFCom2017 papers cover diverse topics in the areas of energy supply, demand and price forecasting.

The guest editorial is HERE. Below is the list of special section papers:
  1. Tao Hong, Jingrui Xie, and Jonathan Black. Global Energy Forecasting Competition 2017: Hierarchical probabilistic load forecasting
  2. Florian Ziel. Quantile regression for the qualifying match of GEFCom2017 probabilistic load forecasting.
  3. I. Dimoulkas, P. Mazidi, and L. Herre. Neural networks for GEFCom2017 probabilistic load forecasting.
  4. Slawek Smyl and N. Grace Hua. Machine learning methods for GEFCom2017 probabilistic load forecasting.
  5. Andrew J. Landgraf. An ensemble approach to GEFCom2017 probabilistic load forecasting.
  6. Cameron Roach. Reconciled boosted models for GEFCom2017 hierarchical probabilistic load forecasting.
  7. Julian de Hoog and Khalid Abdulla. Data visualization and forecast combination for probabilistic load forecasting in GEFCom2017 final match.
  8. Isao Kanda and J.M. Quintana Veguillas. Data preprocessing and quantile regression for probabilistic load forecasting in the GEFCom2017 final match.
  9. Stephen Haben, Georgios Giasemidis, Florian Ziel, and Siddharth Arora. Short term load forecasting and the effect of temperature at the low voltage level.
  10. Jakob W. Messner and Pierre Pinson. Online adaptive lasso estimation in vector autoregressive models for high dimensional wind power forecasting
  11. Dazhi Yang, Elynn Wu, Jan Kleissl. Operational solar forecasting for the real-time market.
  12. Grzegorz Marcjasz, Bartosz Uniejewski, and Rafał Weron. On the importance of the long-term seasonal component in day-ahead electricity price forecasting with NARX neural networks.
  13. Bartosz Uniejewski, Grzegorz Marcjasz, and Rafał Weron. Understanding intraday electricity markets: Variable selection and very short-term price forecasting using LASSO.
  14. Xuerong Li, Wei Shang, and Shouyang Wang. Text-based crude oil price forecasting: A deep learning approach.

Citation 

Tao Hong and Pierre Pinson, "Energy forecasting in the big data world," International Journal of Forecasting, in press, 2019. 

Energy Forecasting in the Big Data World

Tao Hong and Pierre Pinson

Modern information and communication technologies have brought big data to virtually every segment of the energy and utility industries. While forecasting is an important and necessary step in the data-driven decision-making process, the problem of generating better forecasts in the world of big data is an emerging issue and a challenge to both industry and academia. This special section aims to collect top-quality forecasting articles that document cutting-edge research findings and best practices on a wide range of important business problems in the energy industry. Our emphasis is on big data, such as forecasting with high resolution data, the use of high-dimensional processes, forecasting in real-time, and the use of non-traditional data and variables. 

Wednesday, June 5, 2019

SWEET Sessions @ ISF2019

Update 6/21/2019: the SWEET presentations can be downloaded via this Dropbox link. The next ISF will be held at Rio, Brazil, July 5-8, 2020. Look forward to seeing you there!

At the board meeting during the 38th International Symposium on Forecasting (ISF2018), I proposed the idea of developing interest groups or communities within the International Institute of Forecasters (IIF) to better offer a collaborative environment and networking opportunities to forecasting researchers and practitioners. Right after ISF2018, I worked with George Athanasopoulos, Stephan Kolassa and Pam Straud to develop a formal proposal to the IIF Board of Directors. The board approved the launch of two communities at the end of last year. One of them is the Section on Water, Energy and Environment (SWEET).

ISF2019 will be held at Thessaloniki, Greece, June 16 - 19. The conference program committee has dedicated a full 3-day track to SWEET. In total, 43 speakers will cover a wide range of topics in 13 sessions, including gas and electricity demand forecasting, wind and solar forecasting, water demand and hydro generation forecasting, water and air quality forecasting, and energy price forecasting. In addition, we will hold the first SWEET member meeting on Monday June 17, right before the IIF member meeting.

If you are interested in ISF2019, please check out the program schedule. Below is the list of SWEET talks:

Electricity Demand 1: Data Resolution
  1. Forecasting individual electric utility customer hourly loads from AMI data
  2. Development of an end-use load forecasting model for Peninsular Malaysia
  3. Daily peak load forecasting with mixed-frequency input data
Electricity Demand 2: Short Term Load Forecasting
  1. Evaluation of multi-horizon strategies for electricity load forecasting
  2. Zero initialization of modified gated recurrent encoder-decoder network for short term load forecasting
  3. Impact of meteorological variables in short-term electric load forecasting
Electricity Demand 3: Load & Price
  1. Determining the demand elasticity in a wholesale electricity market
  2. Horse and Cart: a scalable electricity load and price forecast model
  3. Temporal hierarchies with autocorrelation for load forecasting
Electricity Demand 4: Statistics vs. Machine Learning
  1. Forecasting time series with multiple seasonal patterns using a long short-term memory neural network methodology
  2. Statistical and machine learning methods combination for improved energy consumption forecasting performance
  3. Probabilistic forecasting of electricity demand using Markov chain and statistical distribution
Electricity Price 1: German Market
  1. Econometric modelling and forecasting of intraday electricity prices
  2. On the importance of cross-border market integration under XBID: evidence from the German intraday market
  3. A generative model for multivariate probabilistic scenario forecasting
Electricity Price 2: Probabilistic Forecasting
  1. Averaging probabilistic forecasts of day-ahead electricity prices across calibration windows
  2. Regularization for quantile regression averaging. A new approach to constructing probabilistic forecasts
  3. Revisiting the jackknife method for construction of prediction intervals – application to electricity market
Electricity Price 3
  1. Forecasting Italian spot electricity prices using random forests and intra-daily market information
  2. Forecasting Northern Italian electricity prices
  3. Application of a SVM-based model for day-ahead electricity price prediction for the single electricity market in Ireland
Electricity Price 4
  1. Day-ahead vs. intraday - forecasting the price spread to maximize economic benefits
  2. Enhancing wind and solar generation forecasts to yield better short-term electricity price predictions
  3. Prediction intervals in high-dimensional regression
Energy
  1. Forecasting algorithm assignment to distribution grid service points in the context of demand response
  2. Modelling uncertainty: probabilistic load forecasting using weather ensemble predictions
  3. Understanding the impacts of distributed PV resources on short-term load forecasting – a comparative study on solar data availability
  4. Access forecasting for safety-critical crew transfers in offshore environments
Environment
  1. A feature-based framework for detecting technical outliers in water-quality data from in situ sensors
  2. Probabilistic forecasting models for NO2 concentrations
  3. Probabilistic forecasting of an air quality index
Oil & Gas
  1. Forecasting oil and natural gas prices with futures and threshold models
  2. Ensemble-based approaches and regularization techniques to enhance natural gas consumption forecasts
  3. A multi-granularity heterogeneous combination approach to crude oil price
  4. Predicting Natural Gas Pipeline Alarms
Water
  1. Forecasting power generation for small hydropower plants using inflow data from neighboring basins
  2. Probabilistic short-term water demand forecasting
  3. When is water consumption extreme?
  4. Forecasting water usage demand in Sydney
Wind & Solar
  1. A comparison of wind speed probabilistic forecast via quantile regression models
  2. Online distributed learning in wind power forecasting
  3. Probabilistic solar power forecasting: long short-term memory network vs. simpler approaches
  4. A non-parametric approach to wind power forecast

If you can't join the conference but want to stay informed about SWEET activities, you can sign up for the SWEET News Letter

Monday, April 22, 2019

Combining Weather Stations for Electric Load Forecasting

10 years ago, I started looking into how weather data quality issues affect load forecast accuracy. Later, I found that using data from multiple weather stations can help improve the load forecasts (see this SAS white paper). I also invented a weather station selection methodology to automatically select weather stations for a given load zone. After joining UNC Charlotte, I wrote an IJF paper with two collaborators to introduce that methodology. Nowadays many utilities are using it to select their weather stations. Because that IJF paper is reproducible, I often use it as an entrance exam for prospective students interested in joining BigDEAL.

During the past few years, I have been using that IJF paper as a homework problem in my Energy Analytics class. I have been challenging the students to improve the weather station selection methodology. Although the method is hard to beat, every year some students can turn in something better. Last year, I decided to work with the students in the class to write two papers, one on selecting weather stations, and the other on combining weather stations. Right after I made that decision, Antonio Bracale and Pasquale De Falco invited me to write a paper related to ensemble forecasting for a special issue they were editing. Weather station combination apparently fits the scope very well. Although I believed the research deserves publication with a higher tier journal, I accepted the invitation to make this paper open access, with the hope that those who are using the old methodology can upgrade to this new one with minimal effort.

The peer review process was fairly enjoyable. The paper was submitted on March 18, 2019. The first decision, which was a major revision, was sent back to us on April 1, with comments from three reviewers. Most of the review comments were constructive. None of them were as nonsense as some of the reviewers I encountered at IEEE transactions. We submitted the revision on April 8. The paper was accepted on April 12. The editorial office sent me the edited version for proofread on April 16. I was presently surprised that their copy editor did some wordsmith for us. I submitted the proofread version on April 20. The final version was published on April 21.

Citation

Masoud Sobhani, Allison Campbell, Saurabh Sangamwar, Changlin Li, and Tao Hong, "Combining weather stations for electric load forecasting," Energies, vol. 12, no. 8, pp. 1510, April 2019. (open access)

Combining Weather Stations for Electric Load Forecasting

Masoud Sobhani, Allison Campbell, Saurabh Sangamwar, Changlin Li, and Tao Hong

Abstract

Weather is a key factor affecting electricity demand. Many load forecasting models rely on weather variables. Weather stations provide point measurements of weather conditions in a service area. Since the load is spread geographically, a single weather station may not sufficiently explain the variations of the load over a vast area. Therefore, a proper combination of multiple weather stations plays a vital role in load forecasting. This paper answers the question: given a number of weather stations, how should they be combined for load forecasting? Simple averaging has been a commonly used and effective method in the literature. In this paper, we compared the performance of seven alternative methods with simple averaging as the benchmark using the data of the Global Energy Forecasting Competition 2012. The results demonstrate that some of the methods outperform the benchmark in combining weather stations. In addition, averaging the forecasts from these methods outperforms most individual methods.

Monday, April 8, 2019

Global Energy Forecasting Competition 2017: Hierarchical Probabilistic Load Forecasting

Check out the winning methodologies and data used in GEFCom2017! If you don't have access to ScienceDirect, you can use the dropbox link below to access the data.

Citation

Tao Hong, Jingrui Xie, and Jonathan Black, "Global Energy Forecasting Competition 2017: Hierarchical Probabilistic Load Forecasting," International Journal of Forecasting, in press. (ScienceDirect; Data)


Global Energy Forecasting Competition 2017: Hierarchical Probabilistic Load Forecasting

Tao Hong, Jingrui Xie, and Jonathan Black

Abstract

The Global Energy Forecasting Competition 2017 (GEFCom2017) attracted more than 300 students and professionals from over 30 countries for solving hierarchical probabilistic load forecasting problems. Of the series of global energy forecasting competitions that have been held, GEFCom2017 is the most challenging one to date: the first one to have a qualifying match, the first one to use hierarchical data with more than two levels, the first one to allow the usage of external data sources, the first one to ask for real-time ex-ante forecasts, and the longest one. This paper introduces the qualifying and final matches of GEFCom2017, summarizes the top-ranked methods, publishes the data used in the competition, and presents several reflections on the competition series and a vision for future energy forecasting competitions.

Thursday, April 4, 2019

Lagged Load Variables in Load Forecasting

This post was triggered by the email below:
I am a regular reader of your blog and website which is an inspiration to me as a forecasting analyst. I just have a very simple question for you, which I don’t understand as a practitioner. I have looked at 10-20 papers and almost every one has a lag variable in it for forecasting electricity demand. But in practice, if you are forecasting for a portfolio or a region and not the whole grid of a country, lag demand is simply not available until weeks or months later. Is this because academia is focused on the theoretical and not the practical, or is it because it focuses on the big picture, total demand and not by region/portfolio? And is there any way round this? You can always feed forecasts for D+1 as a lag into D+2 going forward, but this doesn’t give you a lag for D+0 and D+1.
This is an excellent and frequently asked question, but I don't have a simple answer. 

In practice, if you have lagged load as a variable in the model but don't have its observation for the forecasting period, you have to use the predicted value. 

Take day-ahead load forecasting for example, when forecasting hour ending 10am for tomorrow, we don't have the observation for hour ending 9am. If the model included the lagged load of the preceding hour, we have to predict the load of hour ending 9am first.  In order to make that prediction, we need the load of hour ending 8am, which has to be predicted as well. 

Let's say you are building is a multiple linear regression model, the regression models with lagged dependent variables are called dynamic regression models. 

To implement a dynamic regression model to forecast the period where the observations for the lags are not available, you will have to execute an iterative process to forecast those lags first. 

Now you may want to ask:
Are these dynamic regression models more accurate than the ones without lagged load?
Practically, it depends upon how far ahead you are forecasting and how far back the lagged variables go to. 

If you are using the load of the preceding hour in your model, you should expect some improvement for the next few hours comparing with the models without lagged load variables. The improvement diminishes as the forecast horizon stretches. Beyond 10 hours or so, you may not see any improvement. 

One way to get around this iterative process is to avoid using the load of preceding one or two hours. Instead, we can use the load of the same hour of yesterday. By doing so, you can expect some improvement for the next day or two comparing with the models without lagged load variables. Again, the improvement diminishes as the forecast horizon stretches. For the very short horizon, i.e., one or two hours ahead, the models with the load of the same hour of yesterday typically do not outperform the models with the load of the preceding hour. 

For long term load forecasting, adding lagged load variables doesn't help much but creates issues. 

One is on the interpretability of the model. Because the lagged load variables are highly correlated with the load series itself, most of the load variation is being "explained" by lagged load variables rather than the other explanatory variables such as weather and calendar variables. In other words, we can hardly answer "what if the next year is a hot year" if lagged load variables were in the model. 

Another issue is on the inflation of forecast accuracy. Many people are plugging in actual values of the lagged load when analyzing the long term load forecasting performance, which would result in a very low error. Be careful, this is not ex post forecasting! You should not assume the perfect knowledge of the dependent variable in ex post forecasting. 

To keep the answer short, this is what I have been doing: I use lagged load (see this MPCE paper) when the forecast horizon is less than two or three days, Sometimes I include residual forecasting (see the point forecasting portion of this IJF paper). I don't use lagged load for long term load forecasting. 

Hope this helps!

Tuesday, April 2, 2019

Zehan Xu - Pursuing Perfection

Yesterday (April 1, 2019), Zehan Xu defended his MS thesis Customer Attrition Modeling and Forecasting.

Zehan Xu's MS thesis defense
From left to right: Dr. Linquan Bai, Zehan Xu, Dr. Tao Hong, and Dr. Shaoyu Li


Zehan received his B.S. degree in Industrial and Systems Engineering from Virginia Tech in 2016. He joined our MSEM program in Fall 2017.

During his first semester, I gave a seminary talk about the research opportunities at BigDEAL. He approached me after that, passed the tests I gave him, and officially joined my research group in February 2018.

Knowing his solid math background, I asked him to work on forecasting customer count using survival analysis. The topic was an extension of Jingrui Xie's MS thesis and TSG paper. Since Zehan did not have much background in statistics, he had to teach himself about survival analysis. He quickly figured out that the tools working well on those textbook examples are not optimal for the real-world datasets I gave him. During the past year, he has been refining his work and finally came up with an effective methodology.

Our original plan was to have him graduate at the end of 2018, when I considered the quality of his work exceed a MS thesis level. Nevertheless, he was never satisfied until very recently.

I stopped by my office last Sunday, and saw one of my student Saurabh Sangamwar in the conference room presenting something. Since Saurabh already defended his thesis a month ago, I was a little curious. I went in and found him and another BigDEAL student Yike Li working with Zehan on Zehan's defense rehearsal.

I thought Zehan's defense preparation was done, but apparently he was pursuing that perfection.

His defense was very well done. I was impressed!

While advising him for the thesis research, I found Zehan a great candidate for doctoral research.  He also realized the need and value of advanced education, so he decided to continue pursuing his doctoral degree here at BigDEAL.

Congratulations, Zehan!

Wednesday, February 13, 2019

Saurabh Sangamwar - Nothing is Impossible

Yesterday (February 12, 2019),  Saurabh Sangamwar defended his MS thesis Grouping Calendar Variables for Electric Load Forecasting.

Saurabh Sangamwar's MS thesis defense
From left to right: Dr. Liquan Bai, Saurabh Sangamwar, Dr. Pu Wang, Dr. Tao Hong

Saurabh received his B.Eng. in Mechanical Engineering from K J Somaiya College of Engineering, Mumbai, in 2015. After working in India for two years, he joined our MSEM program in Fall 2017. 

I still remember the scene of our first conversation a year ago, when he expressed his interest in joining BigDEAL.

"learn SAS and get the SAS Base Programmer Certification." I told him the same as what I said to the other students.

"I did." Saurabh said. 

"Then go ahead and get the SAS Advance Programmer Certification." I responded. 

"I've done that too." He said. 

Apparently, he came to me so well prepared, and he was the first student I met this well prepared. 

I admitted him without a blink. 

The topic I gave him is about grouping calendar variables. It took him a while to get the preliminary results. Then I asked him to change a few parameters in his algorithms, and refresh the results. I took him another long while to get the second batch done. I saw him working hard everyday, so I was wondering why it took so long to get the results. During the conversation, I realized that his code is not fully automated. In other words, he had to do a lot of manual work to get the results. I also understood that he did have any programming background until last semester, when he was preparing for the SAS certification exams. 

I'm a professor who likes to pull the students out of their comfort zone. Knowing his weakness, I increased the programming requirements in his master thesis research, so that he can sharpen his programming skills. 

Saurabh did not disappoint me. Over the following few months, he automated his code, picked up parallel computing techniques, and even learned additional languages such as Python and R. Moreover, he is one of the few students took two tough courses from me and got a 4.0 GPA. 

To Saurabh, nothing is impossible. 

Congratulations!

Monday, February 11, 2019

Short-term Industrial Reactive Power Forecasting

Two years ago, I started collaborating with a team of Italian researchers. We had our first joint paper on short-term industrial load forecasting published at the 2017 ISGT-Europe. The complete story is HERE.

Since then, we've continued our collaboration. In this paper, we used the data from the same Italian factory. Now we focus on reactive power forecasting, a rarely touched topic in the load forecasting literature. 

Citation

Antonio Bracale, Guido Carpinelli, Pasquale De Falco, and Tao Hong, "Short-Term Industrial Reactive Power Forecasting," International Journal of Electrical Power & Energy Systems, vol.107, pp 177-185, May 2019 (ScienceDirect)

Short-term Industrial Reactive Power Forecasting

Antonio Bracale, Guido Carpinelli, Pasquale De Falco, and Tao Hong

Abstract

Reactive power forecasting is essential for managing energy systems of factories and industrial plants. However, the scientific community has devoted scant attention to industrial load forecasting, and even less to reactive power forecasting. Many challenges in developing a short-term reactive power forecasting system for factories have rarely been studied. Industrial loads may depend on many factors, such as scheduled processes and work shifts, which are uncommon or unnecessary in classical load forecasting models. Moreover, the features of reactive power are significantly different from active power, so some commonly used variables in classical load forecasting models may become meaningless for forecasting reactive power. In this paper, we develop several models to forecast industrial reactive power. These models are constructed based on two forecasting techniques (e.g., multiple linear regression and support vector regression) and two variable selection methods (e.g., cross validation and least absolute shrinkage and selection operator). In the numerical applications based on real data collected from an Italian factory at both aggregate and individual load levels, the proposed models outperform four benchmark models in short forecast horizons.

Tuesday, December 4, 2018

Leaderboard for BFCom2018 Final Match!!!

The final match of the BigDEAL Forecasting Competition 2018 was on probability daily peak hour forecasting, a very important problem in today's electricity market but new to the academic literature. Even without any monetary prize, all 16 finalists from 5 countries submitted their forecasts. (See the qualifying match leaderboad HERE.)

The figure below shows the leaderboard for BFCom2018 Qualifying Match. The green highlighted ones are in-class students. I also created a naive forecast, which is highlighted in red. 

BigDEAL Forecasting Competition 2018 Final Match Leaderboard

One of my students Zehan Xu, who was auditing the class but got disqualified in the qualifying match, also worked on the final problem and submitted his forecast on time. I included his score on the leaderboard, but marked his ranking as "BR-6", which means bragging right for ranking #6. His ranking does not affect the rankings of the other teams. 

Congratulations to all the BFCom2018 finalists for completing this competition! 

To get updates about the follow-up events, please follow my twitter and/or connect to me on LinkedIn.

Sunday, December 2, 2018

Temperature-based Models vs. Time Series Models

Last week, Spyros Makridakis asked me a question:
I have been reading your energy competition and I cannot find any clear statements about the superiority of explanatory/exogenous variables. Am I wrong? Is there a place where you state the difference in forecasting accuracy between time series and explanatory multivariate forecasting as it relates to the short as well as beyond the fist two or three days (not to mention the long term) that accurate temperature forecasting exist?
Today, Rob Hyndman asked me a similar question, which was routed originally from Spyros.

In fact, this has been quite a debatable topic in load forecasting. The answer is not straightforward. This subject could make a good master's thesis or even a doctoral dissertation. I was going to write a paper about it, but always had something more important or urgent to work on. Recently my research team has done some preliminary work along this direction. While the paper is still under preparation, let me start the discussion with this blog post, as part of the blog series on error analysis in load forecasting.

The literature is not vacant in this area. Various empirical studies have suggested different things.

Some earlier attempts were made by James Taylor. James has written many load forecasting papers. His best known work is on exponential smoothing models.

James' TPWRS2012 paper claimed that
Although weather-based modeling is common, univariate models can be useful when the lead time of interest is less than one day.
In Fig. 9 of the paper that depicted the MAPE values by lead time, the paper stated that
The exponential smoothing methods outperform the weather-based method up to about 5 hours ahead, but beyond this the weather-based method was better. 
Based on this paper, can we conclude that exponential smoothing models are more accurate than the weather-based methods for very short term ex ante load forecasting?

No.

This is my interpretation of the paper:
A world-class expert in exponential smoothing carefully developed several exponential smoothing models. These models generated more accurate forecasts than a U.K. power company's forecasts. 
The "weather-based method" used in that paper was devised by the transmission company in Great Britain using regression models. The paper briefly mentioned how the "weather-based method" worked, but the information was not enough for me to judge how accurate these weather-based models are. I don't know if this U.K. transmission company is using state-of-the-art models.

Some evidence came from recent load forecasting competitions, such as Global Energy Forecasting Competitions, npower forecasting challenges, and BigDEAL Forecasting Competition 2018.

In short, time series models, such as exponential smoothing and ARIMA models, never showed up as a major component of a winning entry in these competitions. On the other hand, regression models with temperature variables are always among the winning models.

In fact, ARIMA showed up in a winning method in GEFCom2014, where my former student Jingrui Xie used four techniques (UCM, ESM, ANN, and ARIMA) to model the residuals of a regression model (see our IJF paper).

Based on these competition results, can we conclude that time series models are not as accurate as regression models?

No.

In GEFCom2012, we let the contestants predict a few missing periods in the history without restricting the contestants to using only the data prior to each missing period. In my GEFCom2012 paper, I briefly mentioned that
This setup may mean that regression or some other data mining techniques have an advantage over some time series forecasting techniques such as ARIMA, which may be part of the reason why we did not receive any reports using the Box–Jenkins approach in the hierarchical load forecasting track.
In GEFCom2012, npower forecasting challenges, and the qualifying match of BFCom2018, actual temperature values were provided for the forecast period. In other words, these competitions were on ex post forecasting. Again, the temperature-based models have an advantage since perfect information of temperature is given for the forecast period.

GEFCom2014 and GEFCom2017 were on ex ante probabilistic forecasting. The temperature-based models dominated the leaderboards. This would be a fair evidence favoring temperature-based models.

For benchmarking purpose, I included two seasonal naive models in my recency effect paper per the request of an anonymous reviewer. Both performed very poorly compared with the other temperature-based models. I commented in the paper:
Seasonal naïve models are used commonly for benchmarking purposes in other industries, such as the retail and manufacturing industries. In load forecasting, the two applications in which seasonal naïve models are most useful are: (1) benchmarking the forecast accuracy for very unpredictable loads, such as household level loads; and (2) comparisons with univariate models. In most other applications, however, the seasonal naïve models and other similar naïve models are not very meaningful, due to the lack of accuracy. 
Here is a quick summary based on the evidence so far:

  • For ex post point load forecasting, evidence favors temperature-based models.
  • For ex ante point load forecasting, no solid evidence favoring either method. 
  • For ex ante probabilistic load forecasting, evidence favors temperature-based models.

I'm not a fan of comparing techniques. In my opinion, it's very difficult to make fair comparisons among techniques. If I were good at ANN but bad at regression, I could build super accurate ANN models than regression models. Using exactly the same technique, two forecasters may build different models with distinct accuracy levels. My fuzzy regression paper offers such an example. In other words, the goodness of a model is largely depending upon the competency of the forecaster. The best way to compare techniques is through forecasting competitions. 

In practice, weather variables is must-have in most load forecasting situations. I'll elaborate this in another blog post.