temporal-gnn

HoloScroll 全息卷轴 ：存在，探索，成为自己。AI 驱动的个人成长科技树：自动聚合你的代码/笔记/推文/健康/日记，生成可解释的技能树并推荐下一步最优行动。愿景：构建属于每个人的成长哲学 HoloScroll 的设计初衷并不仅仅是一款工具，而是一种成长理念的探索与实践。它传递的思想是：成长不应该被片面追求速度或完美，而是一场逐渐构建和完善的旅程，每一步的积累都为下一步铺垫可能性。 这款工具不仅为用户解决了记录割裂和方向迷茫的问题，更试图通过渐进式思想的表达，激发每个人对自我成长的深刻思考。它是一个属于成长探索者的沉稳伙伴，一个帮助用户将点滴积累转化为体系化成就的得力助手，一个可以持续完善的非凡平台。 在 HoloScroll的边界，成长的无限可能正在等待被发现。

S&P 500 equity analysis with temporal graph neural networks, sparse stock graphs, and PyTorch Geometric.

S&P 500 equity analysis with temporal graph neural networks, sparse stock graphs, and PyTorch Geometric.

A Temporal Graph Neural Network platform for corporate stress prediction across the S&P 500. Combines temporal graph attention, LLM-augmented features from SEC filings, and NLP drift detection to identify financial distress signals before traditional models react.

Time-Relaxed Directed GNN for Bitcoin Fraud Detection | 6 Novel Contributions | Production-Ready | E7-A3: 0.5846 PR-AUC (+4.1%) | E9 Fusion: +33.5% | Publication-Ready Research

Deep learning for systemic financial risk: an Unsupervised Graph Autoencoder and a Spatio-Temporal GNN (GCN+LSTM) trained on 26 years of BIS banking network data (2000–2026) to autonomously detect crises and forecast cross-border exposure shifts — no manual features, no crisis labels.

🕐 Uncover Bitcoin fraud with our zero-leakage temporal GNN, designed for effective detection through a systematic, experimental approach.