{"id":"W3154200574","doi":"10.2139/ssrn.3812473","title":"Deep Reinforcement Learning for Algorithmic Trading","year":2021,"lang":"en","type":"article","venue":"SSRN Electronic Journal","topic":"Stock Market Forecasting Methods","field":"Decision Sciences","cited_by":14,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Toronto","funders":"","keywords":"Reinforcement learning; Foreign exchange market; Arbitrage; Currency; Computer science; Markov chain; Q-learning; Foreign exchange; Artificial intelligence; Exchange rate; Machine learning; Economics; Financial economics; Monetary economics","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.02138522,0.0001830692,0.0003397347,0.0002709595,0.0006911596,0.0003647472,0.0006292361,0.00008692207,0.0003335907],"category_scores_gemma":[0.01424122,0.00015188,0.0003444772,0.0007291537,0.00003845579,0.0002890684,0.00009288643,0.001781382,0.00003312445],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008819553,"about_ca_system_score_gemma":0.002362011,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000004463893,"about_ca_topic_score_gemma":0.00007814282,"domain_scores_codex":[0.994556,0.0006041233,0.000808633,0.0004379704,0.001130992,0.002462353],"domain_scores_gemma":[0.9960073,0.002523065,0.0004486613,0.0003031756,0.0005711509,0.0001466591],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.00008522607,0.00002389625,0.001127065,0.000003433833,0.0001320803,0.00001305177,0.0003959364,0.01035949,0.0008030009,0.04379667,0.0003823208,0.9428778],"study_design_scores_gemma":[0.001083024,0.0003987078,0.0002712029,0.0000214829,0.00004162516,0.0018177,0.004637221,0.2180431,0.0006577239,0.7543697,0.01840226,0.0002562744],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01359494,0.001626672,0.9787598,0.0006536477,0.0007208327,0.0001482112,1.756411e-7,0.00003404705,0.004461667],"genre_scores_gemma":[0.9377288,0.0004092966,0.03588711,0.0001445883,0.0008629816,0.00002150293,0.000003468426,0.00004457148,0.02489767],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9428727,"threshold_uncertainty_score":0.9940622,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06310182009114429,"score_gpt":0.3757627060276205,"score_spread":0.3126608859364762,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}