Tag
论文提出 ASE（Attribution-Guided Semantic Evolution）模块化框架，通过条件归因反馈引导语言模型构建与修订文本因子，并将其与数值预测器融合，在 TESS 数据集和 CSI 300 股票排序任务上取得改进。
财新报道：香港出生、剑桥毕业的量化工程师何卓峰（音）被控从前雇主Headlands窃取高频交易策略AP1，并与Tower合作在中国期货市场部署，2025年1月在洛杉矶被捕，纽约陪审团审判已开始，面临最高30年刑期。
Third Moment Research 使用 OpenAI 的 Codex、GPT-6 Astra、GPT-6 Sol 和 GPT-6.1 Sol 构建自我改进的 AI 系统，可同时运行 10,000 个智能体进行持续 36-48 小时的研究，并通过自动合成场景修复智能体弱点（例如用 Codex 将回测时间从 3 小时缩短至 3-5 分钟）。
Jane Street intern Kavish explored whether autoregressive diffusion models can generate realistic market data event streams, finding that fully continuous diffusion struggles with the jagged, spiky distributions of real order book data and that flow matching outperformed DDPM.
Quantitative Finance with OCaml is a comprehensive technical book/guide covering derivatives pricing, risk management, credit modeling, algorithmic trading, and HFT infrastructure using OCaml, with production-quality code, honest mathematical derivations, and coverage of modern OCaml 5 features (domains, effects, OxCaml extensions).
该研究评估了 Llama-3 8B、Qwen-2.5 14B 和 Llama-3.3 70B 等 LLM 在从 SEC 10-K 文件中提取财务数据时的表现，发现将模型规模与文档结构复杂度对齐可在零样本设置下获得高准确率，从而缓解量化金融数据集中超过 70% 公司面临的缺失数据偏差问题。
Researchers from Allora Foundation and Liverpool John Moores University introduce the CZAR (Composite Zero-Agnostic Return) loss, a convex, asymmetric, piecewise-quadratic objective tailored for financial log-return prediction that prevents models from collapsing to zero-forecast behavior, and show improved directional accuracy on BTC intraday returns with LightGBM.
FinRL is an open-source framework for financial reinforcement learning, with about 16,000 stars on GitHub, supporting training and backtesting for trading in stocks, cryptocurrencies, etc., and pointing to the next-generation version FinRL-X.
The tweet shares free code from the book 'Machine Learning in Finance: From Theory to Practice', which integrates machine learning with quantitative finance for data modeling and financial decision-making.
Horizon is an AI-powered tool that allows traders to describe trading strategies in plain language, automatically generating backtests with detailed metrics for iterative testing and improvement.
The article discusses the impact of Stephen Boyd's 2006 mathematics textbook on convex optimization, used by Google, Renaissance Technologies, and others, and promotes his free Stanford course EE364A available on YouTube.
The paper explains how removing market beta and factor exposure reveals true trading signals, emphasizing rigorous stress testing to avoid overfitting and false confidence in backtests.
The paper reports an improvement in Information Coefficient from 0.0613 to 0.0843 using an agent-guided research process, compared to a GRU baseline, with simulated results on held-out data from 2021 to 2025.
Man Group has open-sourced ArcticDB, a high-performance, serverless DataFrame database designed for Python data science and quantitative finance scenarios, supporting billions of rows of time-series data, S3/LMDB/Azure storage, time travel and snapshots, and C++ engine acceleration.
Introduces tvscreener, a Python library that can directly fetch TradingView stock screener data, supports multiple markets and 13,000+ technical indicators, returns Pandas DataFrame, and is convenient for quantitative strategy use.
Promotional post about Horizon, an AI-driven trading strategy platform that lets users backtest and deploy strategies in plain English, citing a professor managing $200M with AI agents and returning 56% last year.
Kirk Borne highlights the 3rd edition of 'Machine Learning for Trading' by Stefan Jansen, a comprehensive 826-page book on building AI-driven trading systems, and announces an accompanying live workshop on August 15.
This preprint introduces AlphaSchema, a framework that constructs and explores a structured space of trading semantics for LLM-based alpha mining, decoupling semantic exploration from factor implementation. Experiments on the Chinese stock market show that it discovers factor pools with strong predictive and portfolio performance.
This paper studies LLMs for parent-order execution in algorithmic trading, introducing PACE, a hierarchical framework that outperforms traditional baselines on Shenzhen Stock Exchange data and suggests LLMs can complement human traders.
Awesome AI in Finance is an open-source curated resource list covering papers, tools, datasets, and trading frameworks for AI and quantitative finance, organized into categories like agents, research, strategies, data, and trading stacks.