數據分析工程師(特徵工程與訊號最佳化)
Feature Engineering & Signal Optimization
【工作內容】
·聚焦於開發進階型、非傳統或跨市場橫截面因子,強化現有因子庫的多樣性與深度。
·探索特徵工程方法(如多層次指標、衍生品因子、主題型或產業鏈因子),補足簡單因子難以捕捉的市場特徵。
·應用統計學、機器學習與遺傳編程等方法,進行因子組合優化與非線性信號挖掘。
·分析並驗證因子在不同市場狀態、風格輪動下的穩健性與解釋力,協助提升策略多樣性。
·與現有因子庫工程師協作,推動因子驗證自動化、因子庫結構優化及新因子納入流程。
·支援模型團隊,將高階因子嵌入預測架構,並參與ablation test、因子貢獻度分析。
Job Responsibilities
·Focus on developing advanced, non-traditional, or cross-market factors to enrich the existing factor library.
·Explore feature engineering (e.g., multi-level indicators, derivatives-based, thematic, or supply chain factors) to capture nuanced market signals.
·Apply statistical, machine learning, and genetic programming techniques for factor combination optimization and non-linear signal discovery.
·Analyze and validate factor robustness and explanatory power across varying market regimes and style rotations.
·Collaborate with current factor library engineers to automate validation, optimize library structure, and streamline new factor integration.
·Support the modeling team by embedding advanced factors into predictive architectures and participating in ablation and contribution analysis.
【我們希望你具備】
·熟悉因子投資理論、統計分析與特徵工程,有進階因子設計或非傳統信號挖掘經驗。
·精通 Python 資料分析與機器學習工具(如 pandas、scikit-learn、XGBoost、statsmodels)。
·能獨立設計因子驗證流程,解釋因子與標的之關聯性,並評估其市場適應性。
·有跨市場、產業或衍生品數據應用經驗者優先。
·具備與資料工程、因子庫團隊協作經驗,能推動因子自動化驗證與流程優化。
·曾用 Qlib、AITrader 或自建因子測試架構者尤佳。
Qualifications
·Solid understanding of factor investing, statistical analysis, and advanced feature engineering; experience in developing non-traditional or complex factors.
·Proficient in Python data analysis and machine learning frameworks (e.g., pandas, scikit-learn, XGBoost, statsmodels).
·Capable of independently designing validation pipelines and interpreting factor-target relationships and market adaptability.
·Experience with cross-market, industry, or derivatives data is a plus.
·Demonstrated ability to collaborate with data engineering and factor library teams to drive automation and workflow optimization.
·Experience with Qlib, AITrader, or custom factor research platforms is highly desirable.
【合作方式]
·兼職/專案制合作,彈性時程,以高階因子研發與驗證成果為主。
·薪資依任務複雜度、因子創新性與驗證精度評估。
·表現優異者可長期參與核心策略設計,協助建立高階因子資產庫。
Collaboration & Compensation
Part-time or project-based collaboration, flexible scheduling, focusing on advanced factor development and validation.
Compensation based on task complexity, factor innovation, and validation rigor.
Outstanding contributors may join long-term in core strategy and advanced factor library development.