Distributional neural networks for electricity price forecasting

Marcjasz, Grzegorz; Narajewski, Michał; Weron, Rafał; Ziel, Florian

doi:10.1016/j.eneco.2023.106843

Cited by 22 publications

(3 citation statements)

References 55 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Accounting for the uncertainties of the price forecasts allows market participants to better evaluate the risks associated with different trading activities, e.g., managing a power portfolio [41]. Recent works in electricity price forecasting evaluate the economic implications of probabilistic forecasts.…”

Section: Economic Evaluation Of Electricity Price Forecastsmentioning

confidence: 99%

“…Recent works in electricity price forecasting evaluate the economic implications of probabilistic forecasts. For example, authors in [36,41,42] quantify the profitability of probabilistic price forecasts in the German day-ahead market by implementing a trading strategy that looks to optimally allocate bids to charge and discharge a battery storage system.…”

Section: Economic Evaluation Of Electricity Price Forecastsmentioning

confidence: 99%

See 1 more Smart Citation

Forecasting the Occurrence of Electricity Price Spikes: A Statistical-Economic Investigation Study

Zamudio López,

Zareipour,

Quashie

2024

Forecasting

View full text Add to dashboard Cite

This research proposes an investigative experiment employing binary classification for short-term electricity price spike forecasting. Numerical definitions for price spikes are derived from economic and statistical thresholds. The predictive task employs two tree-based machine learning classifiers and a deterministic point forecaster; a statistical regression model. Hyperparameters for the tree-based classifiers are optimized for statistical performance based on recall, precision, and F1-score. The deterministic forecaster is adapted from the literature on electricity price forecasting for the classification task. Additionally, one tree-based model prioritizes interpretability, generating decision rules that are subsequently utilized to produce price spike forecasts. For all models, we evaluate the final statistical and economic predictive performance. The interpretable model is analyzed for the trade-off between performance and interpretability. Numerical results highlight the significance of complementing statistical performance with economic assessment in electricity price spike forecasting. All experiments utilize data from Alberta’s electricity market.

show abstract

Section: Economic Evaluation Of Electricity Price Forecastsmentioning

confidence: 99%

Section: Economic Evaluation Of Electricity Price Forecastsmentioning

confidence: 99%

Forecasting the Occurrence of Electricity Price Spikes: A Statistical-Economic Investigation Study

Zamudio López,

Zareipour,

Quashie

2024

Forecasting

View full text Add to dashboard Cite

show abstract

“…In order to benchmark the proposed model's performance, two benchmark models were implemented: First, an autoregressive process with exogenous regressors and seasonal effects (Hamilton, 2020) (SARX) was implemented as a linear point-forecast model that is widely used in time-series analysis. Second, a distributional ANN (Marcjasz et al, 2023) was used as black-box benchmark model trading computational efficiency and interpretability for the ability to learn highly non-linear relationships from the data (Yu et al, 2019). For all benchmarks, the lags L = {1, …, 6} were chosen in order to condition the prediction on a history window equivalent to the proposed model.…”

Section: Benchmarksmentioning

confidence: 99%

Probabilistic Intraday Wastewater Treatment Plant Inflow Forecast Utilizing Rain Forecast Data and Sewer Network Sensor Data

Sonnenschein,

Ziel

2023

Water Resources Research

Self Cite

View full text Add to dashboard Cite

Forecasting of wastewater treatment plant inflow dynamics constitutes an enabler technology for wastewater treatment process optimization using model predictive control. However, accurate inflow prediction is still challenging, especially for strong rainfall events, where complex system dynamics and missing information on future rainfall represent limiting factors. We propose a seasonal probabilistic time series model for modeling the short‐term wastewater inflow accurately while providing quantification of forecast uncertainty. To ensure suitability for practical implementation, the unconstrained parameters of the predictive distribution are modeled as linear functions of the input variables in the framework of Generalized Additive Models for Location Scale and Shape. Non‐linear effects are approximated by Rectified Linear Units, accounting for buffering within the sewer network and flow‐dependent catchment response time. In addition to water level measurements from within the sewer network and rain rate measurements, rain forecasts are incorporated as exogenous regressors, where historical rain forecasts are used for model calibration. The model performance is evaluated on historical data from a German wastewater treatment plant using deterministic and probabilistic scoring rules. We benchmark against an autoregressive time series model and a long short‐term memory artificial neural network. Our results show that the proposed model unites the benefits of high prediction accuracy of the neural network and enhanced intelligibility of the autoregressive model, but accurate real‐time rain forecasts are mandatory for successful real‐world implementation.

show abstract

A feature selection method based on Shapley values robust for concept shift in regression

Sebastián,

González-Guillén

2024

Neural Comput & Applic

View full text Add to dashboard Cite

Feature selection is one of the most relevant processes in any methodology for creating a statistical learning model. Usually, existing algorithms establish some criterion to select the most influential variables, discarding those that do not contribute to the model with any relevant information. This methodology makes sense in a static situation where the joint distribution of the data does not vary over time. However, when dealing with real data, it is common to encounter the problem of the dataset shift and, specifically, changes in the relationships between variables (concept shift). In this case, the influence of a variable cannot be the only indicator of its quality as a regressor of the model, since the relationship learned in the training phase may not correspond to the current situation. In tackling this problem, our approach establishes a direct relationship between the Shapley values and prediction errors, operating at a more local level to effectively detect the individual biases introduced by each variable. The proposed methodology is evaluated through various examples, including synthetic scenarios mimicking sudden and incremental shift situations, as well as two real-world cases characterized by concept shifts. Additionally, we perform three analyses of standard situations to assess the algorithm’s robustness in the absence of shifts. The results demonstrate that our proposed algorithm significantly outperforms state-of-the-art feature selection methods in concept shift scenarios, while matching the performance of existing methodologies in static situations.

show abstract

Distributional neural networks for electricity price forecasting

Cited by 22 publications

References 55 publications

Forecasting the Occurrence of Electricity Price Spikes: A Statistical-Economic Investigation Study

Forecasting the Occurrence of Electricity Price Spikes: A Statistical-Economic Investigation Study

Probabilistic Intraday Wastewater Treatment Plant Inflow Forecast Utilizing Rain Forecast Data and Sewer Network Sensor Data

A feature selection method based on Shapley values robust for concept shift in regression

Contact Info

Product

Resources

About