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
The document summarizes a research study that applies hidden Markov models to identify regimes for individual assets and adjust portfolio exposure in response to price trends. It describes a global allocation analysis spanning January 2004 through December…
The strategy estimates sector return states with a Markov chain, which models each state as dependent on the preceding state. It uses the resulting conditional mean and variance estimates as inputs to a Black–Litterman portfolio process, treating the…
This example applies a linear support vector machine to classify each new bar’s next close-to-close movement as upward, downward, or relatively flat. It builds training examples from historical bar open and close prices, labels each example according to…
This report overview describes how artificial intelligence may be applied across investment management, including portfolio research, investment decisions, trading, risk control, product development, and client marketing. It highlights information…
This weekly report surveys Chinese equity-selection factors built from intraday trading behavior. Examples include return skewness, downside-volatility share, buying interest and large-order net buying after the open, late-session trading share, average…
The document compares TradingAgents, presented as a framework for multi-agent financial research, with Lumibot, described as a Python trading framework that can place agent workflows inside a strategy lifecycle. It outlines how research and debate agents can…
The document outlines a financial forum presentation on the use of artificial intelligence in quantitative investing. Its listed examples include a random forest approach to market timing and hedging, a pattern matching method for industry rotation, and a…
This introductory tutorial explains machine learning concepts before focusing on linear regression. It defines features and labels, distinguishes supervised from unsupervised learning, and outlines classification, regression, clustering, and parameterized…
StockRanker is a tree-based learning-to-rank tool for ordering stocks from supplied features. Its listwise approach optimizes the quality of a stock list’s ranking, and the guide connects the method to the RankNet, LambdaRank, and LambdaMART line of work.…
This guidance explains why AI-generated strategy logic, code, reviews, backtests, and parameter recommendations should be treated as research aids rather than final trading decisions. Outputs may contain errors or omissions, vary across models or settings,…
FreqAI is presented as an open source framework for training machine learning models to forecast market targets from user defined indicators. Users supply features and future looking labels; the framework trains a model for each listed trading pair and…
This function creates a synthetic binary classification dataset for studying feature importance and redundancy in an asset-management machine-learning context. It first generates informative and noise features, then constructs additional redundant features…
This document describes the first BigQuant strategy development interest group, scheduled for May through July 2024. Its purpose is to help learners build strategy templates and demonstrations, reproduce common methods, and improve their implementation…
Condor is presented as an open-source framework for building autonomous trading agents. Its architecture separates an agent layer, where large language models observe conditions and make decisions through an Observe, Orient, Decide, Act loop, from an…
This configuration describes a Chinese equity forecasting experiment using Qlib’s Alpha158 features and an LSTM model. It selects 20 features, applies robust feature normalization and missing-value filling, and ranks labels cross-sectionally. The prediction…
This tutorial explains how convolutional neural networks can improve on a simpler model for classifying MNIST handwritten digits. It reports accuracy of about 99.2% for the CNN, compared with 91% for the earlier model, but gives no experimental setup or…
This overview introduces artificial intelligence as the design of systems that make decisions or predictions, including methods such as machine learning and probabilistic reasoning. It explains why these methods can help with complex tasks involving many…
This document describes a long-or-short crossover strategy that searches pairings among 70 moving average and filter types for a chosen market and timeframe. The selected series use a common period, and an additional exponential smoothing step is intended to…
This article explains how alternative internet and transaction data can supplement traditional equity factors. A typical fund process first defines a stock universe, filters it using industry and financial measures, scores candidates with financial and…
The report summary describes a two-stage stock selection process intended to enhance a fundamental factor model. First, the fundamental model creates an initial stock universe; a deep learning model then screens that set again. The stated rationale is to…
This page summarizes a research paper on using intraday price and volume data to build enhancement signals. Its outline moves from converting high-frequency observations into lower-frequency features, to using genetic programming to discover price-volume…
This introductory guide describes a supervised machine-learning workflow for quantitative equity strategies. It begins by defining the stock universe and target, such as a future return, volatility, or ranking, then splitting historical observations by time…
This forum question compares portfolio holding ratios under two execution schedules: buying with one TWAP setting and selling with another, versus buying at the open and selling at the close. It provides a trading function for a ranked stock strategy that…
The document reviews feature selection as a preprocessing step for machine learning stock selection. It distinguishes filter, wrapper, and embedded approaches, and describes selection criteria such as variance, entropy, F statistics, mutual information,…