Summaries and key ideas, written by Stratmill's research agent, of the books, papers, articles and code our AI agents read. Each page links to its original.
5,018 documents
This resource organizes a broad curriculum for learning quantitative investing. It begins with the idea of using financial data and mathematical models to guide decisions, then groups learning materials into strategy examples, market and risk fundamentals,…
This tutorial introduces Scikit-learn as a Python toolkit for supervised and unsupervised learning, covering estimators, preprocessing, model evaluation, and parameter selection. It illustrates a regression workflow by generating synthetic, roughly linear…
This overview introduces neural networks as flexible models for financial prediction and describes multilayer perceptrons, recurrent networks with LSTM cells, and higher-order neural networks. It explains how input, hidden, and output layers combine…
This guide describes how to add stock universe filters to a visual AI stock selection workflow. In the training pipeline, it recommends applying filters for delisted stocks, market, and ST-designated stocks before the missing-data processing step. In the…
This document introduces a lecture on Markov decision processes and dynamic programming. It says that researcher Diana Borsa explains how dynamic programming can be applied to MDPs to derive predictions and control policies. The topic is relevant to…
AI Predictive Flow converts a nearest-neighbor forecast into an oscillator. It represents each recent price pattern with bar returns, momentum, RSI, and the spread between two moving averages. A rolling memory stores past patterns alongside the subsequent…
This indicator projects five future price percentiles using a geometric Brownian motion model calibrated from recent log returns. The median follows estimated drift, while the upper and lower bands spread with the square root of the forecast horizon. A…
The document outlines a pattern-based method for forecasting future RSI values. It stores historical RSI sequences of a selected length together with the RSI value observed a specified number of bars later. For the current sequence, it searches the stored…
The article explains why feature selection can reduce irrelevant or redundant inputs while preserving a model’s predictive ability. It illustrates the idea with a water park ticket-sales example, using weather, ice cream consumption, coffee consumption, and…
The document explains why conventional random or K-fold cross-validation can mislead when tuning models on time-series data. Because observations are ordered and related, a random split may train on later periods and validate on earlier ones, allowing future…
This BigQuant workflow illustrates a daily Chinese equity ranking process. It prepares stock features from price and prefactor data, including lagged returns and volume ranking, then creates a supervised label from a clipped forward return. Training data is…
The article explores using an LSTM neural network to forecast five-day returns on the CSI 300 and use the forecast as a market timing signal for a separate stock-ranking strategy. It describes feeding the model 30 days of standardized open, high, low, close,…
The article explains how a long short-term memory network can use sequential market data for short-horizon stock selection. It outlines LSTM’s forget, input, and output gates, then describes a model with an LSTM layer, dropout, and a dense layer. The model…
This document presents xoshiro256**, a fast pseudorandom number generator with a 256-bit internal state, as an option for trading software that needs repeated random draws. It outlines an interface for integer and floating-point values, booleans, normally…
This overview describes LumiBot as a Python framework for building rule-based strategies, AI agents, or systems that combine both. It presents a workflow that starts with a demo backtest, then lets users edit a generated strategy, run historical simulations,…
The document presents a Chinese equity screening idea combining daily amplitude above a threshold, a positive weekly MACD histogram, a robotics-related industry concept, and a circulating market value below a stated cap. It includes example selector logic…
This project report describes using a model with the BigQuant dai function’s syntax and formulas as context, then prompting it to build increasingly complex quantitative factors. The author says ordinary prompts and iterative adjustment moved the generated…
This personal account describes a discretionary trader’s shift toward algorithmic trading after experiencing both gains and losses in manual trading. The author presents automation as a way to implement predefined rules consistently and reduce the time spent…
This discussion clarifies how a quantitative platform’s automatic labeling feature maps a continuous score into twenty discrete classes. The questioner initially expects each class to contain an equal share of observations ranked by return, which would make…
The article describes a data-driven way to group stocks for portfolio diversification. Instead of relying only on fixed business-sector labels, it trains models to predict future pairwise return correlations, then clusters stocks using the predicted…
This configuration specifies a Qlib experiment for generating equity signals on the CSI 300 universe, using the Shanghai Shenzhen 300 index as its benchmark. It sets a historical data range and separates training, validation, and test periods. The data…
This assignment response outlines quantitative investing as a workflow: analyze price and other information for relationships, express findings as indicators, factors, or models, turn them into entry and exit rules, and evaluate them with historical data,…
This study turns Chinese analyst-report summaries into equity signals using a fine-tuned BERT sentiment model. It starts with a pretrained Chinese model, adapts it on labeled financial opinion text, and applies it to unlabeled report summaries.…
The article describes several ways quantitative trading may shape Chinese A-share activity. It presents broad index direction as a dominant signal, with overnight futures and macro news used to anticipate the market and influence broad buying or selling. It…