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 Pine library collects statistical measures for assessing time-series forecasts. Its point-forecast metrics include absolute, squared, root mean squared, normalized root mean squared, and percentage errors. It also offers interval-based calculations that…
This indicator presents logistic regression as a binary classification method that estimates event probabilities through a sigmoid function. Its implementation updates model weights with gradient descent over a configurable lookback window on each new bar,…
This long-only strategy trains a small XGBoost model on a rolling sample and uses its estimated probability of an upward move to guide entries and exits. It builds six inputs from RSI, price distance from an EMA, rate of change, normalized ATR, relative…
This indicator classifies market conditions as bullish, bearish, or neutral and estimates how often one state transitions to another over a rolling lookback. Its regime engine uses z-score-normalized price returns by default, or combines up to three…
This indicator combines six moving-average families—EMA, Hull, WMA, DEMA, linear-regression, and Kaufman adaptive averages—into adaptive trend signals. It evaluates variants across selectable lengths and assigns weights based on generated equity. A…
This indicator identifies candidate support and resistance from confirmed pivot highs and lows. For each pivot, it encodes RSI relative to 50 and candle body size relative to ATR as binary features, then applies a logistic function to estimate whether the…
This library implements discrete, linear convolution for two numeric arrays: a signal and a filter of weights. For each output position, it multiplies aligned signal and filter elements and sums the products that overlap. The output includes the full,…
FlyTV maps selected circuits from a measured adult fruit fly connectome into a simulated neural network driven by market candles. Its modeled sensory input emphasizes price declines, rises, and bar range. Looming detectors can trigger an escape response that…
This indicator augments a conventional RSI with a k-nearest-neighbors calculation. It describes comparing current conditions with historical observations using features such as RSI, momentum, volatility, regression slope, and price momentum. Outcomes from…
This indicator combines trend, mean reversion, and momentum scores into a bounded oscillator. Exponential moving averages, RSI, ATR, and momentum measures provide the inputs, while a Best-Worst Method model converts user comparisons of the three factors into…
This strategy combines a conventional ATR-based SuperTrend with a nearest-neighbor filter and trailing take-profit and stop-loss levels. It keeps historical SuperTrend values that match the current direction and fall within distance bounds derived from…
This indicator builds eight features from RSI, including its level, slope, acceleration, and volatility-related measures, then compares current conditions with a stored history of labeled outcomes. A weighted Lorentzian distance selects similar bars, whose…
This indicator provides multiple configurable Nadaraya-Watson kernel regression lines for analyzing price at different speeds. Each line can use a rational quadratic, Gaussian, periodic, or locally periodic kernel, with settings for lookback window,…
This indicator adapts a simple k-nearest-neighbors idea to a price series. It compares adjacent observations across a configurable window using absolute price differences, selects values according to a distance-based rule, and averages the resulting…
The article describes Sahara AI as a blockchain-based platform for creating, registering, licensing, and monetizing AI assets such as datasets, models, and agents. It outlines four layers: user applications, blockchain transactions and identity, data storage…
The article describes Manadia (UMXM) as a proposed infrastructure protocol connecting external data, AI agent decisions, and blockchain settlement. Its architecture has three parts: a data layer that aggregates and challenges inputs, agents that process…
The document describes Bittensor as a decentralized network where miners contribute AI models or computing resources and validators score their outputs. It says rewards in TAO are directed toward stronger-performing subnets, linking participant incentives to…
The document describes Cros as a blockchain-based system intended to connect game developers and advertisers. Its proposed workflow includes campaign management and analytics, programmatic ad auctions, APIs and SDKs, and payments in either fiat or…
The article presents five equity candidates—SentinelOne, Guidewire Software, Procore Technologies, ExlService Holdings, and Exelixis—and summarizes the investment rationale for each. Its comparisons focus on revenue and recurring-revenue growth, operating…
The document reviews Nvidia’s fiscal 2025 third-quarter results and connects the company’s reported growth to broader enthusiasm for AI-related stocks. It highlights revenue and earnings growth, the concentration of sales in data center products, demand from…
The document presents Wisdomise AI as a financial platform built around five products: an AI assistant for market intelligence and task execution, a catalog of financial products, a strategy builder, an intelligence library, and a personal dashboard. The…
The document presents Massa as a layer-one blockchain designed around parallel block production and autonomous smart contracts. Its parallel architecture is compared with sequential block creation: multiple blocks can be produced and connected at once, while…
The article introduces Allora as a decentralized network that combines predictions from independently operated machine learning models. It describes workers submitting forecasts, reputers evaluating them after outcomes are known, and topic coordinators…
The document presents Nesa as a Layer 1 blockchain for private, verifiable AI inference. Users and applications submit model requests to distributed nodes, which process and return results. An Artificial Intelligence Terminal is described as a standardized…