{"paper_id":"2503.13544","title":"Decision by Supervised Learning with Deep Ensembles: A Practical Framework for Robust Portfolio Optimization","authors":["Juhyeong Kim","Sungyoon Choi","Youngbin Lee","Yejin Kim","Yongmin Choi","Yongjae Lee"],"paper_date":"2025-03-16","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2503.13544v7","page_url":"https://thequant.space/flowcharts/decision-by-supervised-learning-with-deep-ensembles-a-pract/","scores":{"empirical_rigor":9,"math_complexity":3,"quadrant":"Street Traders"},"proposes":{"status":"available","statement":"The paper proposes Decision by Supervised Learning (DSL), a practical framework that uses supervised learning and deep ensembles for robust portfolio optimization.","quote":"We propose Decision by Supervised Learning (DSL), a practical framework for robust portfolio optimization.","summary":"We propose Decision by Supervised Learning (DSL), a practical framework for robust portfolio optimization. DSL reframes portfolio construction as a supervised learning problem: models are trained to predict optimal portfolio weights, using cross-entropy loss and portfolios constructed by maximizing…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"Risk‑free rate is set to zero in backtesting.","type":"market","stated":"explicit","quote":"we set the risk-free rate to zero within our backtesting environment for simplicity and transparency."},{"assumption":"A two‑day trading lag is applied to all trades.","type":"execution","stated":"explicit","quote":"our framework incorporates a two-day trading lag"},{"assumption":"Transaction costs of 0.4% per trade (both buy and sell) are applied.","type":"execution","stated":"explicit","quote":"applies realistic transaction costs of 0.4% per trade, both for buying and selling."},{"assumption":"The model is restricted to long‑only portfolios.","type":"model","stated":"explicit","quote":"Our framework currently employs cross-entropy loss and is designed for long-only portfolios."},{"assumption":"Cross‑entropy loss is used as a convex surrogate for portfolio weight prediction.","type":"model","stated":"explicit","quote":"training models to predict optimal portfolio weights directly using convex surrogate losses—such as cross-entropy—rather than non-convex financial metrics."},{"assumption":"Ensembles of independently trained neural networks reduce predictive variance.","type":"statistical","stated":"implied","quote":"Deep Ensemble methods, substantially reducing variance in portfolio allocations."}]},"datasets":{"status":"available","items":[{"name":"Eight investment universes (Large Cap, Range‑bounded, High Dividend, Utility Sector, NASDA","provider":null,"kind":"market-prices","access":"unknown","period":"January 2010 to May 2025","universe":"Various manually and objectively selected stock sets","quote":"We evaluate our framework and all baseline methods across eight distinct investment universes"}],"known":[],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"daily"},"transaction_costs":{"status":"available","modeled":"yes","headline_net_of_costs":null,"detail":"0.4% per trade, both for buying and selling","quote":"applies realistic transaction costs of 0.4% per trade, both for buying and selling.","source":"front_matter","rule":"R3","candidates":{"front_matter":"yes","evidence_card":"yes","evidence_card_quoted":true},"card_detail":"0.4% per trade, both for buying and selling","card_quote":"applies realistic transaction costs of 0.4% per trade, both for buying and selling.","meaning":"costs are part of the reported headline results"},"results":{"status":"available","evaluation_type":"trading-pnl","findings":[{"finding":"DSL consistently outperforms machine‑learning baselines (PFL and E2E) and classical strategies, achieving higher cumulative returns, Sharpe and Sortino ratios across most universes.","kind":"performance","quote":"DSL outperforms both Machine Learning baselines (PFL and E2E) as well as classical strategies (equal-weight, value-weight and mSSRM), achieving higher cumulative returns and better Sharpe and Sortino ratios."},{"finding":"Larger ensemble sizes improve performance: cumulative returns, Sharpe and Sortino ratios all increase as ensemble size grows from 1 to 64.","kind":"performance","quote":"As ensemble size increases from 1 to 64, cumulative returns steadily improve"},{"finding":"DSL achieves the best performance across all tested objectives in stable universes (Large‑Cap, High‑Dividend) and remains competitive in dynamic rolling universes, especially S&P 500.","kind":"performance","quote":"DSL achieves the best performance across all tested objectives, demonstrating its effectiveness when the stock universe is stable"}],"reported_performance":null,"out_of_sample":{"method":"walk-forward","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":13,"citations":{"cited_by_in_corpus":0,"by_independent_authors":0,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":0,"unclassified":0},"independent_papers":[],"scope":"Citations between papers in this corpus only; citations from outside it are not counted."}},"implementation":{"status":"available","public":true,"code_url":"https://github.com/DSLwDE/DSLwDE","note":null},"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":"2026-11-14T17:42:37Z"},"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/"}