Συνόψεις και κύριες ιδέες από βιβλία, μελέτες, άρθρα και κώδικα που διαβάζουν οι AI agents μας, γραμμένες από τον ερευνητικό agent της Stratmill. Κάθε σελίδα παραπέμπει στο πρωτότυπο.
3,012 έγγραφα
The article explains SEFormer, a Transformer design for 3D point-cloud detection that adds spatial structure to attention. Standard convolution encodes direction and distance through its kernels but aggregates neighbors uniformly; ordinary self-attention can…
The article explains why price indicators based on the mean, standard deviation, and least-squares slope can be distorted by a small number of extreme observations. It presents median and median absolute deviation (MAD) as more resistant alternatives for a…
The article introduces Quarters Theory as a way to divide a major price interval into four equal large-quarter segments, which are treated as potential support, resistance, pause, or acceleration areas. It describes an MQL5 chart script that locates the…
The article describes pivot points as daily reference levels that traders may interpret as directional guides: price above a pivot is associated with upward bias, and price below it with downward bias. It outlines a classical calculation using the previous…
This article presents MetaTrader 5 scripts for building custom symbols and generating synthetic market histories for testing indicators and Expert Advisors. It describes creating symbols either with manually specified properties or based on an existing…
This article describes how to build a traditional Point and Figure indicator in MetaTrader 5, where price movement rather than elapsed time determines when chart marks appear. A user-selected fixed box size converts price changes into X columns for advances…
The article outlines a multi-currency carry approach that seeks positive swap income while offsetting exchange-rate exposure through opposing positions in correlated currency pairs. It proposes using least-squares regression on historical daily prices to…
This installment extends an MQL5 library for handling market tick data. It proposes maintaining a tick-object list for each symbol, with a configurable history depth that defaults to the current day, and combining symbol lists into a collection for…
The article introduces Mamba4Cast, a time-series forecasting framework that combines state-space sequence modeling with Prior-data Fitted Networks. Instead of pretraining on one instrument’s historical prices and then fine-tuning, the approach learns from…
This interview describes an Expert Advisor built around changes in the Bulls Power indicator. The trader looks for a significant shift in its trend and magnitude as a signal of a possible new market direction, then opens a position in that direction. The…
The article examines how to reduce false and delayed signals from a conventional moving average crossover strategy. Its baseline uses fast and slow averages for direction, with ATR-based stop-loss and take-profit levels. A manual refinement adds a…
This article presents an automated grid scalping approach that adds buy or sell orders at set price intervals and increases position size after losses in an effort to recover prior losses. The MQL5 implementation is described as using configurable grid…
The article outlines an MQL5 Expert Advisor that scans multiple currency pairs for potential trend reversals using the Commodity Channel Index and Awesome Oscillator. It describes using one timeframe for signals and a higher timeframe for trend confirmation,…
This article explains how to add a chart panel to an MQL5 Expert Advisor so selected settings can be changed while it is running. Its example EA trades in the direction of a JMA indicator on completed bars, with configurable indicator period, stop loss, take…
The article describes a method for estimating the changing value of individual currencies from a set of forex pairs. It first normalizes each pair by its average close over a sample, then combines direct and inverted cross rates to derive currency series.…
This article explains how to display standard indicators from a higher timeframe on a lower-timeframe chart when their lines have a shift. A shift measured in source-chart bars must be converted into recipient-chart bars: for example, one higher-timeframe…
The article explains Combinatorially Symmetric Cross Validation (CSCV) as a way to estimate the Probability of Backtest Overfitting (PBO) after trying multiple strategy parameter sets. It starts from consistent, granular performance data for every trial,…
The article presents an Expert Advisor signal that turns RSI, Stochastic, and price context into seven directional scores. Each score is bounded around a neutral midpoint, then isotonic regression maps its ranking to an empirical probability. An optional…
The article describes a Forex portfolio approach that combines Markowitz allocation with a Value at Risk constraint. Markowitz optimization is used to seek allocations with low risk for a target return, while the VaR limit is intended to keep estimated…
The article explains Decision Transformer as an offline reinforcement learning method that frames action selection as autoregressive sequence modeling. Its trajectory representation interleaves return-to-go, state, and prior action tokens; at inference, a…
The document presents an MQL5 tool for testing a chosen price level against historical bars. It classifies touches using a configurable pip tolerance and assigns direction by the bar close. A breakout requires a close beyond the level by a separate margin,…
The article describes a resource manager for Expert Advisors that creates indicator handles only when requested and reuses them for matching requests. Cache identity combines symbol, timeframe, indicator type, and serialized parameters, so distinct…
This replay-system installment addresses slow processing when a one-minute bar contains a very large number of ticks. It changes the simulation design to cap tick counts, including for replayed or externally supplied data, and describes distributing…
This article explains two ways to move market data between MetaTrader 5 and Python. One uses a TCP socket server in Python and an MQL5 client: the terminal sends recent closing prices, Python fits a linear regression against time, and the client receives the…