Knowledge library

This notebook compares predictive models for ranking NASDAQ-100 stocks by their next 15-minute return using intraday microstructure information such as spreads, depth imbalance, signed volume, and price impact. It selects one representative prediction set…

This notebook derives Kelly sizing for a binary wager by maximizing expected log wealth, then illustrates how stake size changes the distribution of simulated wealth paths. It extends the framework to continuous returns and multi-asset portfolios, where the…

This evaluation screens option-surface, price-derived, and conditional-volatility features against forward equity returns. For each session it computes a cross-sectional rank correlation between feature values and subsequent returns, then examines the time…

This utility describes how to validate a migration that changes backtest identifiers without rerunning the underlying computations. It scans database schemas to discover columns containing registered addresses, including embedded references in text…

This analysis assesses whether signals selected in multiple trading case studies appear dependable enough for further use. It compares validation information coefficients (IC) with holdout evidence, examines stability and reproducibility, and contrasts…

This notebook explains instrumented principal component analysis (IPCA) for an equity panel with option-surface features. Unlike ordinary PCA, which estimates each stock's exposures from its return history, IPCA models exposures as a shared linear function…

This notebook generates predictions for a previously untouched ETF holdout window using the configuration selected on validation data. It rebuilds the training specification to end one label horizon before the window opens, refits the model, preserves the…

This notebook outlines a pipeline for extracting supplier, customer, and competitor relationships from company 10-K filings and storing them as a Neo4j knowledge graph. A local language model proposes structured subject-predicate-object triples, after which…

This document describes registry checks for deciding whether to reuse or rerun model training and backtests. Runs are identified by hashes derived from their specifications; a run is reusable only when its expected artifacts are present. Missing pieces mark…

This notebook demonstrates fine-tuning a transformer for financial named entity recognition, turning text spans into structured organization, person, money, date, and percentage fields. It introduces BIO boundary labels and explains how word-level…

This notebook studies gradient boosting for cross-sectional prediction of CME futures returns. It varies tree capacity through leaf-count profiles and compares squared-error, absolute-error, and Huber objectives, which differ in how strongly extreme…

This notebook explains three ways to turn minute-level NASDAQ-100 data into model-based features: a rolling HAR regression for variance forecasts and forecast errors, a Fourier transform for recent volume patterns, and depth-two path signatures for the order…

This notebook compares four ways to retrieve passages from a corpus of company filing text: BM25, dense embeddings, reciprocal rank fusion (RRF), and a cross-encoder reranker. RRF combines ranked lists rather than raw scores, so it can join lexical and…

This document presents an end-to-end forecasting architecture that combines a research panel, an aggregate forecast, an adversarial debate, and a supervisor. The supervisor identifies disagreements, searches for clarifying evidence, and proposes a…

This notebook organizes cross-case-study comparisons of LightGBM models, using linear models and TabM as reference families where available. It evaluates configurations across tree sizes, regression losses, checkpoints, and classification objectives. The…

This notebook compares LightGBM with an MLP, TabM, and, when available, TabPFN for predicting ETF forward returns from tabular features. Models are evaluated on shared chronological walk-forward folds, with rank information coefficient and training time as…

This notebook compares ordinary least squares with Ridge, LASSO, and Elastic Net for predicting 21-day forward returns across ETFs. The models use the same declared walk-forward folds, with a purge gap matched to the label horizon. Each fold clips feature…

The document examines the ADIA Lab causal discovery challenge as supervised inference within a known synthetic data-generating family. The task assigns variables to eight roles around a treatment and outcome, using many labeled dataset and graph pairs. It…

This module builds a mean forecast ensemble from registered prediction sets, with particular attention to reproducibility and avoiding selection bias. It resolves qualifying model configurations from the registry rather than maintaining a hand-written member…

This notebook explains a local linear trend state-space model for extracting financial features from price levels. The hidden state contains a price level and its slope; observations are noisy closes. At each step, the filter predicts the state, measures the…

This notebook separates continuous price variation from discrete jumps in intraday equity returns. It estimates daily realized variance from squared log returns and uses bipower variation as a jump-robust estimate of the continuous component; their…

This notebook uses principal component analysis on a liquid subset of US equities to extract latent factors and form long-short eigenportfolios. It explains how standardized PCA vectors are converted to raw-return weights and normalized by gross exposure,…

This notebook expresses a four-phase forecasting workflow using plain Python orchestration, CrewAI, and LangGraph. It compares where state is held, how failures are inspected or recovered, what dependencies each approach adds, and how tracing works. The…

This notebook examines linear-model regularization for a universe of currency pairs whose returns are linked because each pair is a quote between shared currencies. It distinguishes feature collinearity, which ridge, lasso, and elastic net can address, from…