Modern ML for Trading
The full gradient-boosting-to-neural-operator toolkit — ensembles, AutoML, probabilistic forecasting, distillation and pruning for latency, plus physics-informed architectures like Neural ODEs and Fourier operators.
- 01
Aug 25, 2026 #machine-learning用於報酬方向的 XGBoost：類別不平衡與決策閾值
報酬方向分類器是不平衡問題，而 0.5 是錯誤的決策閾值。比較加密貨幣資料上的 scale_pos_weight、焦點損失與精確率約束閾值最佳化，以及 XGBoost、LightGBM 和 CatBoost 之間的工程差異。
- 02
Aug 3, 2026 #machine-learning整合方法：結合弱學習者以獲得穩健的 Alpha
本部落格未涵蓋有關交易整合的三件事：為什麼協方差項（而不是模型計數）在整合誤差中占主導地位，堆疊如何作為具有折疊元特徵的訊號組合層工作，以及跨多個訊號的周轉淨額是否真正降低了執行成本。
- 03
Jul 30, 2026 #AutoML用於系統化交易流程的 AutoML
自動特徵生成（tsfresh、Featuretools）、具預算感知的模型搜尋（FLAML 的節省成本最佳化器），以及 WorldQuant 公式化 alpha 工廠——本部落格搜尋與過度擬合系列尚未涵蓋的研究流程部分，以及該系列自身結果對它們的啟示。
- 04
Aug 15, 2026 #deep-learningMulti-Task Learning for Simultaneous Price, Volume, and Volatility Prediction
Does jointly predicting return, volume, and volatility actually help? Measuring loss-balancing schemes and diagnosing negative transfer through gradient cosine similarity — with a classical baseline and purged walk-forward folds.
- 05
Feb 19, 2026 #演算法交易交易機器人保護中的異常檢測：從Z-Score到Transformer
哪些異常檢測方法在加密貨幣演算法交易中真正有效，如何構建級聯防護架構，以及為何這是演算法交易不可或缺的基礎。
- 06
Aug 20, 2026 #forecastingScoring Probabilistic Forecasts: CRPS, PIT Calibration, and DeepAR
How to evaluate a predictive distribution honestly — CRPS as a proper scoring rule, the PIT histogram as a calibration diagnostic, and DeepAR sampling in GluonTS.
- 07
Aug 22, 2026 #HFT逐筆資料模型中的不規則時間：連續時間編碼與純位置嵌入
餵入逐筆資料的序列模型仍假設間隔規則。本文介紹三種告訴 Transformer 某筆成交實際發生時間的方法——可學習時間尺度的連續編碼、ODE-RNN 潛在狀態，以及把 delta_t 作為一般特徵——並以消融實驗決定它們之間的取捨。
- 08
Aug 4, 2026 #deep-learning一整天的內容能勝過十分鐘的內容嗎？ Flash 注意力與序列長度問題
Flash Attention 讓 23,400 步驟的交易上下文無需計算——平鋪、線上 softmax 和 N^2 d^2 / M 的 IO 界限。更長的上下文是否使模型更好是一個單獨的問題，它必須被測量，而不是斷言。
- 09
Aug 11, 2026 #model-compressionKnowledge Distillation: Compressing Trading Models for Low-Latency Deployment
The blog's standing answer to the accuracy-vs-latency tension is a two-stage fast/slow split. Distillation is a different answer: train one small model to mimic the ensemble. The KD loss, temperature, born-again nets, early exits for a variable latency budget, and the distill-to-FPGA pipeline — plus the measurements that would decide whether it beats the two-stage split.
- 10
Aug 14, 2026 #model-compressionModel Pruning for Low-Latency Trading Inference
Magnitude and structured pruning, the Lottery Ticket Hypothesis, movement pruning, distillation and 2:4 sparsity — the methods behind shrinking a trading model, and what still has to be measured before any of it ships.
- 11
Aug 6, 2026 #bayesian非參數價格建模的高斯過程
金融時間序列的核設計——母體粗糙度、局部週期成分、光譜混合——以及作為不需要驗證集的正則化器的邊際似然。加上在任何東西可以交易之前仍然需要衡量的誠實清單。
- 12
Jul 31, 2026 #bayesian認識與對立：衡量收益模型不知道的東西
本博客的每個位置 sizing 規則都把不確定性坍縮成單一純量。MC Dropout 和深度集成將其拆分為模型無知與市場噪聲——這兩者值得不同的 position size。
- 13
Aug 16, 2026 #deep-learningNeural ODEs: Does Continuous Time Beat a Delta-Time Feature?
Neural ODEs, Neural SDEs and continuous normalizing flows for irregularly-sampled market data — and the one ablation that decides whether continuous dynamics are worth their solver cost.
- 14
Aug 8, 2026 #deep-learningHamiltonian Neural Networks: Does a Financial System Conserve Anything?
Hamiltonian Neural Networks are provably stable — but stability is worthless if the conserved quantity does not exist. Learning a scalar H via autograd, symplectic integration, and the falsification test that decides whether a financial (q, p) pair is canonical at all.
- 15
Aug 5, 2026 #deep-learning基於偏微分方程的金融建模的傅立葉神經算子
算子學習映射整個函數空間，而不是點——傅立葉神經算子如何在頻率空間中參數化 PDE 解算子，為選擇權定價購買什麼，以及它的哪些主張仍然需要在真實硬體上測量。