{"paper_id":"2506.08026","title":"TIP-Search: Time-Predictable Inference Scheduling for Market Prediction under Uncertain Load","authors":["Xibai Wang"],"paper_date":"2025-05-30","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2506.08026v2","page_url":"https://thequant.space/flowcharts/tip-search-time-predictable-inference-scheduling-for-market/","scores":{"empirical_rigor":7.5,"math_complexity":6,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes TIP-Search, a deadline-aware inference scheduling framework that dynamically selects a deep learning model per task to maximize predictive accuracy while meeting strict latency deadlines.","quote":"This paper proposes TIP-Search, a time-predictable inference scheduling framework for real-time market prediction under uncertain workloads.","summary":"This paper proposes TIP-Search, a time-predictable inference scheduling framework for real-time market prediction under uncertain workloads. Motivated by the strict latency demands in high-frequency financial systems, TIP-Search dynamically selects a deep learning model from a heterogeneous pool…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"Multiple pre‑trained models, each tailored to a specific domain, are available for selection.","type":"model","stated":"implied","quote":"The system relies on the availability of multiple pre-trained models, each tailored to a specific domain or data distribution."},{"assumption":"Model generalization profiles remain stationary over time.","type":"model","stated":"explicit","quote":"TIP-Search currently assumes stationarity in model generalization profiles over time"},{"assumption":"Each inference task must be completed within a hard 10 ms deadline.","type":"execution","stated":"implied","quote":"runtime inference is triggered every 5ms, with a strict 10ms deadline per task."},{"assumption":"All models are deployed on a single GPU for inference.","type":"execution","stated":"implied","quote":"These models are deployed on a single GPU"},{"assumption":"Offline profiling of latency and generalization performance accurately reflects online behavior.","type":"model","stated":"implied","quote":"Our approach profiles latency and generalization performance offline"},{"assumption":"Task arrivals follow realistic bursty patterns as simulated in the experiments.","type":"market","stated":"implied","quote":"To simulate realistic online inference scenarios, we construct a heterogeneous workload using three distinct financial limit order book (LOB) datasets."}]},"datasets":{"status":"available","items":[{"name":"FI-2010","provider":null,"kind":"order-book","access":"public","period":"ten days of Nasdaq LOB data","universe":null,"quote":"comprises ten days of Nasdaq LOB data annotated with five‑class mid‑price movements.","known_id":"fi-2010"},{"name":"Binance BTC/USDT","provider":"Binance","kind":"order-book","access":"public","period":"2025/05/10 snapshot","universe":null,"quote":"The Binance BTC/USDT dataset contains a 2025/05/10 snapshot of raw cryptocurrency trade events labeled for three‑class price direction.","known_id":"binance"},{"name":"LOBSTER AAPL","provider":"LOBSTER","kind":"order-book","access":"public","period":"ten trading days","universe":"Apple Inc.","quote":"The LOBSTER AAPL dataset provides high‑frequency order book data for Apple Inc. over ten trading days, representing a more stable equity market.","known_id":"lobster"}],"known":[{"id":"fi-2010","name":"FI-2010","provider":"Ntakaris et al.","kind":"order-book","access":"public","quote":"For instance, a convolutional model trained on high-frequency data from Binance may offer low-latency predictions but perform poorly when evaluated on Nasdaq-based FI-2010 data.","mentions":10,"used_in":2},{"id":"binance","name":"Binance","provider":"Binance","kind":"market-prices","access":"public","quote":"For instance, a convolutional model trained on high-frequency data from Binance may offer low-latency predictions but perform poorly when evaluated on Nasdaq-based FI-2010 data.","mentions":9,"used_in":2}],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"l2-book"},"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":"TIP-Search improves predictive accuracy by up to 8.5 % over static baselines while achieving 100 % deadline satisfaction.","kind":"performance","quote":"Our results highlight the effectiveness of TIP-Search in robust low‑latency financial inference under uncertainty."},{"finding":"TIP-Search consistently achieves higher accuracy than all three baselines (fixed‑model, random, and source‑aware).","kind":"predictive-accuracy","quote":"TIP-Search consistently achieves higher accuracy than all three baselines."},{"finding":"Model selection aligns with latent data characteristics even without explicit source labels during execution.","kind":"other","quote":"The observed routing patterns confirm that model selection in TIP-Search aligns with latent data characteristics even without access to explicit source labels during execution."},{"finding":"Classical real‑time scheduling policies (EDF, LLF) yield lower accuracy than TIP-Search under the same deadline constraints.","kind":"performance","quote":"The resulting accuracy under deadline constraints was inferior to TIP-Search, suggesting that incorporating domain‑specific accuracy estimation provides meaningful gains over purely temporal heuristics."},{"finding":"Static selection strategies either violate latency guarantees or suffer from model‑data mismatch, leading to lower overall performance.","kind":"negative","quote":"strategies often lead to suboptimal performance: either violating latency guarantees or compromising predictive accuracy due to model–data mismatch."}],"reported_performance":null,"out_of_sample":{"method":"holdout","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":36,"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":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":"2026-11-14T17:36:56Z"},"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/"}