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.
21,023 documents
This overview explains how neural networks combine neuron types and connection patterns, then surveys architectures including feed-forward networks, Hopfield networks, Boltzmann machines, autoencoders, recurrent networks, convolutional networks, and others.…
This strategy uses a fitted copula and marginal cumulative distribution functions to estimate conditional probabilities for two assets. During a formation period, the model is trained on historical prices. As new prices arrive, their marginal distributions…
The document presents a market-cycle indicator based on the Discrete Fourier Transform. Its stated motivation is that fixed lookback periods in common technical indicators may become poorly aligned when the market’s underlying cycle changes. The proposed…
The Wave Segregation Index (WSI) is presented as a histogram oscillator built from the Commodity Channel Index, the Average Directional Index, and typical price. Its calculation multiplies CCI by typical price and ADX, then divides the product by 1,000.…
This overview surveys a platform’s machine learning tools across classification, regression, ranking, and clustering. It names representative methods such as support vector machines, neural networks, gradient boosting, random forests, nearest neighbors, and…
This document describes an indicator that displays price data as candlesticks after processing the relevant time series with a smoothing algorithm. It presents this format as a potentially more informative way to inspect price behavior in some analytical…
This Chinese stock-selection proposal combines three filters: daily amplitude above 1, a top-five rank by the day’s opening-auction amount, and large-order net volume above 0.05 for at least three consecutive days. The author interprets persistent positive…
This module describes a candidate-selection process for pairs trading based on dimensionality reduction and clustering. It starts from a panel of asset prices, converts prices to returns, standardizes them, and applies principal component analysis to create…
This tutorial introduces finite difference methods for approximating derivatives in a parabolic partial differential equation, using the heat equation as its example. It motivates the approach for quantitative finance by noting that the Black-Scholes…
The document explains how to inspect trained parameters in a Keras model that combines an LSTM layer, an auxiliary input, and several dense layers. It distinguishes trainable variables from the arrays returned by the model’s weight retrieval method, then…
This tutorial explains how to retrieve one field’s value from the day another field reaches its maximum within a rolling window. Its example seeks the trading volume on the day of the highest price over the preceding 20 days. The approach uses `m_imax` to…
This short forum exchange addresses an error in a regression-evaluation workflow that reports a missing label. The response says the problem cannot be diagnosed precisely without details about the task, then recommends first defining what quantity the model…
The Tail_Bar indicator compares the upper and lower shadows of each price bar and plots their difference as a histogram around a zero line. The description presents a longer lower shadow relative to the upper shadow as one side of the comparison and says the…
This report reviews quality-focused Smart Beta investing in overseas and Chinese markets. It describes the approach as an index strategy built from company financial data to capture underlying business quality, then compares product development and index…
The document describes a variant of the stochastic oscillator that uses a step chart to filter small changes. A user sets the step size as a percentage of the stochastic value; changes below that threshold are treated as insignificant, which is intended to…
The document describes a trend-following indicator that adjusts its Average True Range multiplier using Z-score normalization. The stated aim is to widen its trend boundary during volatility spikes and tighten it in steadier conditions, addressing the lag…
The Double Smoothed Stochastic, attributed here to William Blau and Walter Bressert, adapts the stochastic oscillator by applying exponential smoothing twice. The provided calculation first smooths a stochastic reading, rescales that smoothed series within…
The article examines whether dividend announcements convey information about future company earnings. It argues that prior studies may have missed this link by grouping earnings by fiscal year, which can treat post-announcement earnings as already known.…
This indicator profiles five configurable intraday sessions and summarizes their historical behavior for a chosen market. For each session, it records range, volume share, whether it set the day’s high or low, directional closes and continuation, fair value…
This article explains several ways to limit the effect of extreme observations in financial data, where outliers can distort derived features and model training. It covers percentile clipping, which caps values outside chosen quantiles, and a…
SumRSI is an indicator that averages RSI oscillator values calculated across a user-defined range of periods. Its inputs set the starting and ending lookback lengths, the applied price series, and a horizontal bar shift. The supplied defaults span periods…
This research notebook outlines a simple test of whether stocks with stronger prior returns tend to earn higher returns next day. It measures momentum from prices over a twelve-month window while omitting the latest month, then computes next-day returns. The…
This report compares ways to measure similarity among funds, including cosine similarity, Euclidean distance, and return correlations. It argues that the choice of measure should follow the intended meaning of “similar”: resemblance in holdings structure,…
This report compares ways to measure similarity among funds, including cosine similarity, Euclidean distance, and return correlations. It argues that the choice of measure should follow the intended meaning of “similar”: resemblance in holdings structure,…