{"paper_id":"2404.03968","title":"Regularization for electricity price forecasting","authors":["Bartosz Uniejewski"],"paper_date":"2024-04-05","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2404.03968v1","page_url":"https://thequant.space/flowcharts/regularization-for-electricity-price-forecasting/","scores":{"empirical_rigor":8.5,"math_complexity":3,"quadrant":"Street Traders"},"proposes":{"status":"available","statement":"The paper proposes a fully automated approach to select all regularization parameters based on cross‑validation for electricity price forecasting.","quote":"Finally, we propose a fully automated approach to select all the regularization parameters based on CV.","summary":"The most commonly used form of regularization typically involves defining the penalty function as a L1 or L2 norm. However, numerous alternative approaches remain untested in practical applications. In this study, we apply ten different penalty functions to predict electricity prices and evaluate…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"A parsimonious autoregressive (ARX) model structure is used for forecasting.","type":"model","stated":"explicit","quote":"as the first underlying model, we propose a parsimonious autoregressive structure, which is a well-established model in the EPF literature"},{"assumption":"A parameter‑rich fARX model is employed as an alternative forecasting structure.","type":"model","stated":"explicit","quote":"the parameter‑rich fARX model with cross‑validation as the parameter selection method produces the most accurate forecasts"},{"assumption":"Cross‑validation is assumed to be a suitable method for tuning the regularization parameter λ.","type":"statistical","stated":"explicit","quote":"we employ cross‑validation (CV) or Bayesian information criteria (BIC) to select the value of the tuning parameter (λ) and compare their performance"},{"assumption":"The German EPEX SPOT and Spanish OMIE markets are representative electricity markets for the study.","type":"market","stated":"explicit","quote":"We assess the performance of those regularization types using two datasets from the German EPEX SPOT and Iberian OMIE markets"},{"assumption":"The data series are assumed to be hourly and cover a seven‑year period.","type":"data","stated":"explicit","quote":"This empirical study utilized datasets from two distinct European markets, each covering a period of seven years"},{"assumption":"Ordinary least squares (OLS) forecasts serve as the benchmark for evaluating regularized models.","type":"other","stated":"explicit","quote":"The accuracy of price forecasts obtained with regularized models is compared with the benchmark models estimated with ordinary least squares (OLS)"}]},"datasets":{"status":"available","items":[{"name":"German EPEX SPOT market data","provider":"ENTSO‑E transparency platform","kind":"market-prices","access":"public","period":"seven years (up to 30.9.2018 for DE‑AT‑L","universe":"German day‑ahead electricity market (DE‑AT‑LU / DE‑LU)","quote":"The data were obtained from the transparency platform (https://transparency.entsoe.eu) and consist of five hourly time series"},{"name":"Spanish OMIE market data","provider":"ENTSO‑E transparency platform","kind":"market-prices","access":"public","period":"seven years","universe":"Spanish day‑ahead electricity market (OMIE)","quote":"Data were also collected from the transparency platform and comprise four hourly time series"}],"known":[],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"intraday-bars"},"transaction_costs":{"status":"not_applicable","modeled":"n/a","headline_net_of_costs":null,"detail":null,"quote":null,"source":"front_matter","rule":"R3","candidates":{"front_matter":"n/a","evidence_card":"n/a","evidence_card_quoted":false},"meaning":"the paper reports no trading or execution result"},"results":{"status":"available","evaluation_type":"forecast-accuracy","findings":[{"finding":"Across all regularization types, LQ regularization yields the most accurate forecasts on average, outperforming LASSO in every case.","kind":"performance","quote":"Across all the regularization types, the LQ regularization performs best on average. It produces more accurate forecasts than the LASSO‑estimated models in all cases."},{"finding":"Elastic net is the second‑best regularization method, beating LASSO in seven out of eight cases.","kind":"performance","quote":"The second best regularization type in terms of forecast accuracy is the elastic net. In particular, it outperforms LASSO in seven out of eight cases."},{"finding":"For both markets, the parameter‑rich fARX model with cross‑validation for λ selection gives the most accurate forecasts, while BIC selection is not advisable.","kind":"predictive-accuracy","quote":"The results show that for both markets, the parameter‑rich fARX model with cross‑validation as the parameter selection method produces the most accurate forecasts. Conversely, using BIC to select the optimal λ value is not advisable for the majority of regularized models."},{"finding":"OLS estimation remains hard to beat for the parsimonious ARX model, though ridge regression slightly outperforms the OLS benchmark.","kind":"performance","quote":"OLS estimation is hard to beat for the parsimonious ARX model, but ridge regression slightly outperforms the benchmark."},{"finding":"Cross‑validation‑based automated parameter selection achieves accuracy comparable to the best ex‑post selected models without requiring expert knowledge.","kind":"predictive-accuracy","quote":"The accuracy of such models is similar to the best models selected ex‑post, and at the same time it does not require additional expert knowledge."}],"reported_performance":null,"out_of_sample":{"method":"walk-forward","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":19,"citations":{"cited_by_in_corpus":2,"by_independent_authors":0,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":0,"unclassified":2},"independent_papers":[],"scope":"Citations between papers in this corpus only; citations from outside it are not counted."}},"implementation":{"status":"available","public":false,"code_url":null,"note":"No repository presented by the authors as their own was found in the paper. Absence here is not proof that no code exists."},"replication":{"status":"none_found","card":null,"note":"No replication of this paper has been run by thequant.space yet."},"extraction":{"text_source":"fulltext","model":"openai/gpt-oss-120b","version":1,"extracted_at":"2026-11-14T18:11:38Z"},"provenance":{"claims":"Everything here is what the authors state in the paper. Nothing was re-run or reproduced by the platform.","quotes":"Assumptions, datasets, cost detail and findings were extracted by a language model. Every quote is verbatim paper text. Where an item has `support`, \"full\" means every figure in it appears in the quoted sentences and \"partial\" means a figure is stated elsewhere in the paper; items with a figure found nowhere in the paper were discarded.","datasets_known":"datasets.known lists well-known vendors and public datasets by a stable id, found by a dictionary over the full text where the sentence says the data is used; it is independent of the model and covers only names in the dictionary.","flags":"transaction_costs.modeled, results.evaluation_type and results.out_of_sample.method are platform classifications (structured-decision models); see https://thequant.space/score-guide/#penalty-flags","comparisons":"supports / contradicts / qualifies are a language model's reading of two papers' quoted findings, restricted to closely related papers that share no author. They are pointers to read, not verified replications, and coverage is partial.","implementation":"code_url is a repository the paper itself presents as its own (rule-scored, ambiguous cases judged by a model), checked for liveness when it was recorded."},"docs":"https://thequant.space/agents/"}