This article tackles a common MT5 usability problem: users delete a chart object created by an indicator, and earlier implementations removed the entire program. The proposed approach keeps the indicator running and automatically recreates the missing object. It starts with a minimal fix: listen for object-deletion events and rebuild the object. Testing reveals subtle issues: repeated pointer initialization can leak memory, and OnDeinit can accidentally trigger recreation via queued events. The corrected version guards pointer reuse and disables event handling during shutdown. A more practical refinement stores key object state (position and size) inside the object-managing class. When deletion is detected, the indicator restores the object using saved properties, balancing effort against fidelity. 👉 Read | Calendar | @mql5dev
Открыть в TelegramПосты канала MQL5 Algo Trading
G7 FVG OB Confluence AutoTrader is an MT5 algorithmic trading utility built around strict OB/FVG overlap validation. Trade signals are only considered when an institutional Order Block and an unmitigated 3-bar Fair Value Gap coincide at the same price structure, with configurable overlap tolerance. The engine scans M15 order blocks via impulse displacement (>1.6x ATR) and tracks active imbalances. A 6-factor on-chart score (0–100%) aggregates NY-session USD news filtering, DXY correlation alignment, imbalance status, liquidity sweep detection, order block state, and an overall weighted confidence with score capping on structural conflicts. Execution is handled from a chart HUD with one-click buy/sell, lot sizing, SL/TP edits, auto-exec toggling, breakeven-all, and close-all. Risk controls include daily loss shutdown, daily profit lock, spread filt... 👉 Read | Quotes | @mql5dev
Открыть в TelegramДинамика просмотровСкрыть динамику
Рост просмотров
Часов после публикации
Показать таблицей
| Часов после публикации | Просмотры |
|---|---|
| 0 ч | 0 |
| 2 ч | 2 398 |
| 16 ч | 8 574 |
- Всего сейчас8 574
Smart Money Concepts workflows often require monitoring several timeframes and maintaining consistent zone markup. G7 Institutional SMC Radar is an MT5 indicator focused on XAUUSD scalping and prop-firm style execution, providing automated zone detection and a confluence-style confidence readout for manual decisions. The tool auto-draws bullish/bearish Fair Value Gaps and ICT Order Blocks, extends them forward, and removes zones after mitigation to reduce chart clutter. A heads-up display outputs a 0–100 confidence score based on structural alignment between detected zones. A background DXY correlation module is included to help validate directional bias. Recommended attachment is M15 or H1, then waiting for higher confidence readings (for example, 70%+) before executing manual entries. The package is positioned as an analytical dashboard rather than an a... 👉 Read | Freelance | @mql5dev
Открыть в TelegramДинамика просмотровСкрыть динамику
Рост просмотров
Часов после публикации
Показать таблицей
| Часов после публикации | Просмотры |
|---|---|
| 0 ч | 0 |
| 4 ч | 3 750 |
| 18 ч | 7 850 |
- Через 24 часа7 850
- Всего сейчас7 850
An MLP gate can be trained directly inside an MQL5 Expert Advisor to avoid the Python export workflow and its operational failures: scaler mismatches, feature drift, missing weight files, and tester discrepancies. The design targets a compact 2-class MLP (ReLU + softmax) retrained every N bars on the current symbol/timeframe, using only closed-bar inference (shift=1). Behavior when untrained is controlled via InpMlpFailOpen. Integration keeps the existing pipeline intact: RSI divergence generates direction, HTF filter validates context, and the MLP acts as the final accept/reject step based on P(up)−P(down) exceeding InpMlpThreshold for the candidate direction. Features are local (RSI, ATR, EMA(50), session sin/cos), with labels built from forward movement over InpMlpHorizon without live-edge leakage. Trade-offs: CPU cost in tester/optimizer, seed determi... 👉 Read | Docs | @mql5dev
Открыть в TelegramДинамика просмотровСкрыть динамику
Рост просмотров
Часов после публикации
Показать таблицей
| Часов после публикации | Просмотры |
|---|---|
| 0 ч | 0 |
| 6 ч | 4 309 |
| 20 ч | 7 268 |
- Через 6 часов4 309
- Через 24 часа7 268
- Всего сейчас7 268
G7 Pro Trade Manager & Prop Firm Equity Protector is an on-chart trade management utility for manual trading on MetaTrader 5, focused on fast execution and rule-based risk controls. The interface runs directly on the chart and supports scalping and funded-account workflows. Core functions include one-click BUY/SELL with pre-set SL/TP in pips and an on-chart lot sizer. Automated management covers breakeven moves after a configurable profit threshold, plus trailing stops that activate and follow price by defined distances. For prop-firm style limits, a daily drawdown circuit breaker tracks the day’s starting balance versus live equity. When the configured loss threshold is hit, positions can be force-closed and trading locked until the next day. Emergency actions include breakeven-all, partial close, close-profit only, and close-all. Configuration supports symbol... 👉 Read | Docs | @mql5dev
Открыть в TelegramДинамика просмотровСкрыть динамику
Рост просмотров
Часов после публикации
Показать таблицей
| Часов после публикации | Просмотры |
|---|---|
| 0 ч | 0 |
| 6 ч | 3 509 |
| 20 ч | 5 581 |
- Через 6 часов3 509
- Через 24 часа5 581
- Всего сейчас5 581
Chimp Optimization Algorithm (ChOA) extends GWO by using four leaders (attacker, barrier, chaser, driver) and updating each agent from all four simultaneously. A “social motivation” term adds optional chaotic jumps, but this component is often harmful in continuous search. A continuous BChimp implementation reuses the ChOA core and focuses on coefficient dynamics. The published binary variant can hide issues because sigmoid thresholding collapses scale and sign. A key defect is bias in A: using Ak = f(2*c1gk*rand - 1) yields E[Ak] = f(c1gk - 1), creating systematic drift toward bounds and diagonal motion. Fixing this with Ak = f*c1gk*(2*rand - 1) restores zero-mean steps and stable convergence. The tested setup keeps four-leader elitism, resamples coefficients per coordinate, and drops chaotic replacement for continuous benchmarks. 👉 Read | AppStore | @mql5dev
Открыть в TelegramOrder block detection for XAUUSD and any symbol: the zone is defined as the last opposite candle before a structure break. Last bearish candle before a close above a confirmed swing high, or last bullish candle before a close below a confirmed swing low. The zone can be drawn as full range (high to low) or body only. Leg validation uses an impulse filter: at least one candle body in the leg must exceed a configurable ATR multiple (default ATR 14, multiplier 1.0). This reduces zones created by slow drift. Swings are confirmed only after bars on both sides close, so zones appear with a delay and do not repaint. Zone lifecycle: fresh zones are highlighted, tested zones are dimmed on first return, and zones are removed when a candle closes through them. History scan defaults to 1500 bars, with limits for maximum live bullish/bearish zones and optional re... 👉 Read | NeuroBook | @mql5dev
Открыть в TelegramMost MQL5 monitoring ends at terminal logs and Strategy Tester reports. Once an EA depends on external services, correlating decisions and latency across processes becomes manual and unreliable. A native OpenTelemetry OTLP tracer for MQL5 is implemented as a client-only design. It generates W3C trace IDs/span IDs, produces nanosecond-formatted timestamps, and exports OTLP/HTTP JSON via WebRequest(), with no gRPC or protobuf dependency. Instrumentation is layered: core clock/IDs/sampler, tracer with span stack and RAII scope guard, and an exporter with a bounded ring queue plus retry/backoff. OnTimer() performs blocking WebRequest() calls; spans enqueue without blocking, overwriting the oldest on overflow. An EMA-cross EA and a Python scoring service share trace context via the traceparent header, allowing Jaeger to render one end-to-end timeline per b... 👉 Read | AppStore | @mql5dev
Открыть в TelegramRolling forecast filters in MQL5 can be built around Diebold–Mariano with HAC long-run variance, short-sample correction (HLN), and Clark–West for nested AR benchmarks. A practical implementation scores forecasts only after the h-bar horizon closes, handles overlapping errors, and converts one-sided p-values into a trade gate. Output includes an EA, indicator, and per-bar CSV for replay and audit. Tests on XAUUSD, EURUSD, GBPUSD (H1/M15, h=3, 250-bar rolling window) show all three models (EMA slope, linear regression channel, AR(3) on 300 bars) had higher full-sample RMSE than a zero-return random walk. The random walk was significantly more accurate in most comparisons. Gating reduced exposure sharply; profitability was not established as statistically significant under these settings. 👉 Read | VPS | @mql5dev
Открыть в TelegramДинамика просмотровСкрыть динамику
Рост просмотров
Часов после публикации
Показать таблицей
| Часов после публикации | Просмотры |
|---|---|
| 0 ч | 0 |
| 12 ч | 4 070 |
| 26 ч | 5 206 |
- Через 24 часа5 206
- Всего сейчас5 206
Gradient support is added by replacing a constant ARGB fill with a paint abstraction that can be queried per pixel. The rasterizer keeps the existing edge coverage mask and asks Paint::ColorAt(x,y) when compositing. Paint.mqh defines a minimal base class with one virtual method and a virtual destructor. Solid color becomes the simplest paint implementation, not a renderer special-case. Fill keeps the old uint overload by wrapping it in a reusable solid paint instance. Most gradient logic is isolated outside Raster.mqh. Gradients share a stop list (positions 0..1 plus ARGB) stored as parallel arrays, kept sorted on insertion to avoid per-pixel sorting. Sampling clamps to pad-extend, supports hard edges via duplicate positions, and interpolates alpha along with RGB. 👉 Read | VPS | @mql5dev
Открыть в TelegramДинамика просмотровСкрыть динамику
Рост просмотров
Часов после публикации
Показать таблицей
| Часов после публикации | Просмотры |
|---|---|
| 0 ч | 0 |
| 14 ч | 4 444 |
| 28 ч | 5 412 |
- Через 24 часа5 412
- Всего сейчас5 412
Scalping EAs often look stable in the MT5 Strategy Tester but fail live because execution is shaped by microstructure friction: spread expansion, latency-driven price drift, and order-book slippage. On small targets, these costs can consume nearly all edge and flip expectancy after commission. The article presents a pre-trade gatekeeper (CFrictionGuard) that blocks orders when observable conditions imply poor execution. Instead of a static spread limit, it combines tick-based price velocity, tick arrival rate, and adaptive spread anomaly detection (Welford variance transitioning to EWMA) into a Friction Ratio that compares total costs to projected short-horizon movement. Implementation focuses on deterministic OnTick performance: pre-allocated ring buffers, single-pass statistics, a four-regime state machine (optimal/moderate/locked/spike), and safety ... 👉 Read | Quotes | @mql5dev
Открыть в TelegramДинамика просмотровСкрыть динамику
Рост просмотров
Часов после публикации
Показать таблицей
| Часов после публикации | Просмотры |
|---|---|
| 0 ч | 0 |
| 1 ч | 1 127 |
| 16 ч | 5 144 |
| 30 ч | 6 171 |
- Через 1 час1 127
- Через 24 часа6 171
- Всего сейчас6 171
This article fixes a common MT5 pitfall: creating subwindows and objects from an indicator without properly tying their lifecycle to the indicator instance. MT5 blocks adding the same indicator twice, but changing inputs allows new instances—revealing orphaned subwindows when an indicator is removed. The solution is explicit cleanup using ChartIndicatorDelete, with consistent naming shared across files via an include header to keep maintenance manageable as the project grows. It then hardens behavior with chart events: enable object-deletion notifications, detect CHARTEVENT_OBJECT_DELETE, and react deterministically (remove the indicator that owns the deleted object). Unique, traceable object names plus a simple comparison operator prevent collisions with user-created objects and make ownership checks reliable. 👉 Read | NeuroBook | @mql5dev
Открыть в TelegramДинамика просмотровСкрыть динамику
Рост просмотров
Часов после публикации
Показать таблицей
| Часов после публикации | Просмотры |
|---|---|
| 0 ч | 0 |
| 3 ч | 2 533 |
| 18 ч | 5 709 |
| 32 ч | 6 641 |
- Через 24 часа5 709
- Всего сейчас6 641
Daily rollover often brings a short liquidity gap where bid-ask spreads can expand 5x to 15x within seconds while mid-price remains stable. That behavior can trigger pending Buy Stops, hit tight stops on shorts, or push floating equity below a daily loss limit around midnight. Broker Rollover Spread Logger records rollover spreads tick-by-tick and provides an EA pause signal via iCustom. On attach, it seeds a daytime baseline from up to 240 historical bars outside the rollover window, capping seeding spreads at 50 points to avoid skew. During normal hours the baseline is maintained with an outlier-filtered exponential average. During the configured window, any tick exceeding InpSpikeMultiplier (default 3.0x baseline) and InpMinSpikePts sets Buffer 0 to 1.0 on bar 0. Buffers: 0 guard flag (BLOCK/SAFE), 1 current spread points, 2 peak spread for the active o... 👉 Read | AlgoBook | @mql5dev
Открыть в TelegramДинамика просмотровСкрыть динамику
Рост просмотров
Часов после публикации
Показать таблицей
| Часов после публикации | Просмотры |
|---|---|
| 0 ч | 0 |
| 7 ч | 5 364 |
| 36 ч | 8 253 |
- Через 6 часов5 364
- Всего сейчас8 253
A risk calculator indicator can be used to quantify position risk and adjust exposure by changing stop-loss distance. After loading the indicator, a red stop line is plotted at a default offset of 300 pips from the current price. To change the stop level, double-click the red line to activate selection, then click and hold to drag it to the intended stop-loss price. Once the stop line is placed, the tool updates and displays the corresponding position size (lot value) based on the configured risk parameters. 👉 Read | Forum | @mql5dev
Открыть в TelegramДинамика просмотровСкрыть динамику
Рост просмотров
Часов после публикации
Показать таблицей
| Часов после публикации | Просмотры |
|---|---|
| 0 ч | 0 |
| 9 ч | 6 021 |
| 38 ч | 8 330 |
- Всего сейчас8 330
An MT5 EA implements a two moving average crossover entry on the last closed bar. Default parameters are EMA 20 vs EMA 50 on close, with configurable method and applied price. On a confirmed cross it opens a single position in the cross direction, with optional reverse signals. Trade management supports stop loss and take profit in points, and an option to close on the opposite cross. A breakeven module can move SL to entry plus an offset after a defined profit distance, respecting broker stop levels. Position sizing supports fixed lots or risk-based lots (percent of balance or equity). Risk modes require SL enabled. The EA enforces a max concurrent position count, checks free margin before sending orders, and limits actions to its own symbol/magic positions. Points must be adjusted to symbol digits (e.g., XAUUSD 2 vs 3 decimals). Range conditions will prod... 👉 Read | CodeBase | @mql5dev
Открыть в TelegramДинамика просмотровСкрыть динамику
Рост просмотров
Часов после публикации
Показать таблицей
| Часов после публикации | Просмотры |
|---|---|
| 0 ч | 0 |
| 11 ч | 6 720 |
| 26 ч | 8 057 |
- Через 24 часа8 057
- Всего сейчас8 057
An MT5 Fibonacci retracement indicator for XAUUSD, FX, indices, and any symbol, drawing levels on the last confirmed leg. Swings are confirmed using a configurable swing length, requiring closed bars on the right side before a high/low is accepted. The most recent two confirmed swings define the active leg; direction is derived from which swing came first. Retracements include 23.6, 38.2, 50.0, 61.8, and 78.6, with 61.8 in a separate color and optional 127.2/161.8 extensions. Level prices can be printed beside each line, with configurable style, width, and right-side extension. Options include scanning a fixed bar window or forcing “last completed leg” logic, plus an optional 61.8 touch alert firing once per bar. Three hidden buffers are provided for automation: swing high, swing low, and direction (1/-1), updated on the last bar and readable via iCustom... 👉 Read | Freelance | @mql5dev
Открыть в TelegramДинамика просмотровСкрыть динамику
Рост просмотров
Часов после публикации
Показать таблицей
| Часов после публикации | Просмотры |
|---|---|
| 0 ч | 0 |
| 1 ч | 1 950 |
| 13 ч | 6 869 |
| 28 ч | 7 907 |
- Через 1 час1 950
- Через 24 часа7 907
- Всего сейчас7 907
A session-box indicator for MT5 that marks Asian, London, and New York sessions per day and builds a high-to-low rectangle using broker server time. Session highs and lows are projected to the right as dotted lines for a configurable number of bars, with optional labels showing session name and range in points. Configuration includes calendar days to draw back, per-session enable flags, start/end hours, and support for sessions crossing midnight when end hour is below start hour. Display options cover per-session colors, fill, line extension, font settings, and an object prefix to run multiple instances on one chart. The tool is chart-only: no signals, no trading logic. Objects are rectangles, trend lines, and text, refreshed once per new bar to stay lightweight. Intended use is M5–H1; higher timeframes reduce session detail. Range output is in points,... 👉 Read | Forum | @mql5dev
Открыть в TelegramДинамика просмотровСкрыть динамику
Рост просмотров
Часов после публикации
Показать таблицей
| Часов после публикации | Просмотры |
|---|---|
| 0 ч | 0 |
| 3 ч | 3 415 |
| 15 ч | 6 583 |
| 30 ч | 7 620 |
- Через 24 часа7 620
- Всего сейчас7 620
JAFX Advanced SMC is an SMC/ICT-focused indicator that automates market structure and liquidity markup across multiple timeframes, reducing manual chart work. Performance is prioritized via ATR caching and constrained object creation around the live candle to keep charts responsive on large history loads. Structure logic flags Swing and Internal BOS/CHoCH with optional filtering. Liquidity tools include EQH/EQL detection using an ATR-based threshold, plus HH/HL/LH/LL labeling with Strong/Weak levels aligned to current trend. Premium, Discount, and Equilibrium zones are derived from the last confirmed swing range. Order blocks are generated only after a confirmed break by scanning the leg for the most extreme candle, limiting repaint behavior. OB mitigation supports close-based or wick-based invalidation and removes zones when breached. A dashboard re... 👉 Read | Forum | @mql5dev
Открыть в TelegramДинамика просмотровСкрыть динамику
Рост просмотров
Часов после публикации
Показать таблицей
| Часов после публикации | Просмотры |
|---|---|
| 0 ч | 0 |
| 5 ч | 4 116 |
| 17 ч | 6 503 |
| 32 ч | 7 397 |
- Через 6 часов4 116
- Через 24 часа6 503
- Всего сейчас7 397
Cellular10K v3 replaces a fixed feature topology with a trainable feature graph, letting relationships between 50 market factors adapt online. Instead of hardwiring indicator pairs, a connection matrix strengthens edges when feature combinations lead to correct forecasts and weakens them after errors, including learning negative links when context flips a signal’s meaning. The graph layer performs a single normalized message-passing step over features, keeping values bounded. Training uses a local Hebbian-style update rather than backprop, with stabilizers like weight limits, slow learning, normalization, and optional decay to reduce regime-chasing. This graph feeds context-enriched inputs into the 10,000-cell automaton, while the 10-bar binary predictor maintains its own graph for a different horizon. Diagnostics (density, average weight magnitude... 👉 Read | Freelance | @mql5dev
Открыть в TelegramДинамика просмотровСкрыть динамику
Рост просмотров
Часов после публикации
Показать таблицей
| Часов после публикации | Просмотры |
|---|---|
| 0 ч | 0 |
| 9 ч | 5 102 |
| 36 ч | 7 708 |
- Всего сейчас7 708
Skewness and kurtosis are common in trading stats, but both are numerically fragile. Cubed and fourth-power deviations let a single outlier dominate a rolling window, so kurtosis can swing for data-quality events rather than regime change. L-moments use linear combinations of order statistics via probability-weighted moments, keeping shape ratios bounded and more stable under outliers. tau4 is used as a kurtosis analogue, with lambda2 acting as the scale term; tau3/tau4 are undefined when lambda2 is near zero and must be guarded. An MQL5 implementation is built with native arrays only: LMomentsCore.mqh (PWM to L-moments), ReturnWindowStats.mqh (rolling window plus Pearson g1/g2 baseline), and a script validator that walks CopyRates() once and scores spikes via a confusion matrix using ATR-based forward moves. Across three pair/timeframe tests, Pears... 👉 Read | Forum | @mql5dev
Открыть в TelegramДинамика просмотровСкрыть динамику
Рост просмотров
Часов после публикации
Показать таблицей
| Часов после публикации | Просмотры |
|---|---|
| 0 ч | 0 |
| 11 ч | 4 905 |
| 38 ч | 7 176 |
- Всего сейчас7 176
MetaTrader 5 ML workflows often end up with duplicated artifacts across dev machines and VPS instances. Centralizing models and feature snapshots in S3 is practical, but MQL5 has no AWS SDK and S3 requires SigV4 on every request. A native SigV4 implementation can be built using only CryptEncode() and WebRequest(), avoiding DLL dependencies that are blocked in many sandboxed terminals. This still relies on WebRequest() allow-list and local TLS/network policy. Scope is limited to PutObject, GetObject, and ListObjectsV2, implemented as three files: a signer (canonical request, manual HMAC-SHA256, signing-key derivation), an S3 client wrapper, and a script that uploads ONNX/CSV, downloads them back, and validates SHA-256 round trips. The signed headers set is minimal and excludes session tokens and service-specific headers. 👉 Read | NeuroBook | @mql5dev
Открыть в TelegramДинамика просмотровСкрыть динамику
Рост просмотров
Часов после публикации
Показать таблицей
| Часов после публикации | Просмотры |
|---|---|
| 0 ч | 0 |
| 1 ч | 1 238 |
| 13 ч | 5 100 |
| 25 ч | 6 500 |
- Через 1 час1 238
- Через 24 часа6 500
- Всего сейчас6 500
A running CA with 10,000 MT5 agents hit two quality bottlenecks: 50 inputs treated as unrelated scalars, and a Moore lattice that spreads regime shifts slowly while connecting semantically unrelated cells. The upgrade targets only ExtractFeatures() plus neighbor selection and voting, keeping OnTick() fast and the EA interface unchanged. Features become a 50-node graph: a lightweight diffusion step blends each feature with predefined semantic neighbors and a few small-world links. Implemented via index lists (no matrices, no training), this acts like a spectral filter that damps isolated indicator spikes and reinforces coherent multi-indicator structures. Cells also move from pure geometry to a hybrid network: some local neighbors remain, but others are semantic (same feature group, similar strategy) plus long-range chords. Voting weights add a coherenc... 👉 Read | CodeBase | @mql5dev
Открыть в TelegramДинамика просмотровСкрыть динамику
Рост просмотров
Часов после публикации
Показать таблицей
| Часов после публикации | Просмотры |
|---|---|
| 0 ч | 0 |
| 3 ч | 3 009 |
| 15 ч | 6 053 |
| 27 ч | 7 339 |
- Через 24 часа7 339
- Всего сейчас7 339
Trade refusals in MT5 often slip past logs, even after common “rulebook” checks. The article builds a drop-in preflight gate, CTradePreflight, that validates and optionally repairs trade requests before OrderSend(), with unit tests and a repeatable stress EA. The pipeline separates broker-free rules (volume step/min/max, tick-size grid, stop/freeze constraints) from live-terminal checks (permissions, sessions, volume limits, margin) and finishes with OrderCheck(). Repairs round volume down without increasing risk, and can widen stops only when explicitly allowed. A ledger tracks every refusal reason. A key finding: with tester execution delay, some requests still fail because price shifts between validation and server processing. A configurable delay buffer adds safety by requiring extra distance from market, trading fewer borderline orders to avoi... 👉 Read | VPS | @mql5dev
Открыть в TelegramДинамика просмотровСкрыть динамику
Рост просмотров
Часов после публикации
Показать таблицей
| Часов после публикации | Просмотры |
|---|---|
| 0 ч | 0 |
| 7 ч | 6 415 |
| 31 ч | 9 240 |
- Через 6 часов6 415
- Через 24 часа9 240
- Всего сейчас9 240
Perceptron mechanics are often obscured by terminology. Forward propagation is the weighted sum plus bias followed by an activation function, producing an output passed to other units or exposed to callers. Training adjusts weights and bias. With least squares on linear regression, the identity activation is the consistent choice; other activations change the problem definition. Test scaffolding can isolate training, logging, and convergence checks. Learning rate remains the main stability control. Too high causes oscillation around minima; too low slows convergence. A max-iteration guard prevents non-terminating runs. For gradient-based updates, the key addition is multiplying the backpropagated error term by the derivative of the activation function. This is the practical boundary between forward computation and backprop. 👉 Read | Signals | @mql5dev
Открыть в TelegramДинамика просмотровСкрыть динамику
Рост просмотров
Часов после публикации
Показать таблицей
| Часов после публикации | Просмотры |
|---|---|
| 0 ч | 0 |
| 9 ч | 7 494 |
| 33 ч | 9 674 |
- Всего сейчас9 674
Classic SuperTrend on MT5 is built on fixed ATR period and multiplier. When regime or volatility shifts, those constants produce narrow bands and whipsaws, or wide bands and delayed reversals. Manual retuning and one-shot optimizations both fail because parameters stay static while conditions change. An adaptive design replaces constants with live calibration. A background probe matrix scores 5-bar outcomes, tracks MAE, and updates ATR period, multiplier, SL/TP with deadband control and EWM smoothing. A 2D regime grid (trend/sideways vs normalized volatility) stores decayed statistics per cell and returns stable 3x3 neighborhood recommendations. Signal quality is gated by an accuracy layer: EMA8/EMA21 spread, ADX floor, Keltner squeeze block, candle body confirmation, plus a weighted confidence score. L1 regularization zeroes weak feature weights to ... 👉 Read | CodeBase | @mql5dev
Открыть в TelegramДинамика просмотровСкрыть динамику
Рост просмотров
Часов после публикации
Показать таблицей
| Часов после публикации | Просмотры |
|---|---|
| 0 ч | 0 |
| 11 ч | 7 573 |
| 35 ч | 9 265 |
- Всего сейчас9 265
Perceptron training is moved from least-squares updates to gradient-based updates, keeping the same squared-error loss while changing how weights and bias are adjusted. The gradient is treated as a vector: it indicates steepest increase; using the negative direction reduces loss. Implementation relies on partial derivatives for bias and each weight, rather than a secant-style update often confused with a tangent approach. Code changes focus on accumulating per-sample error, computing a shared term for bias/weights, then scaling weight gradients by the corresponding input value. Bias uses a constant derivative of 1. Training loops until the loss drops below a threshold, applying a learning rate and tracking the maximum component error to avoid premature stopping and uniform weight steps. 👉 Read | Calendar | @mql5dev
Открыть в TelegramДинамика просмотровСкрыть динамику
Рост просмотров
Часов после публикации
Показать таблицей
| Часов после публикации | Просмотры |
|---|---|
| 0 ч | 0 |
| 11 ч | 6 224 |
| 35 ч | 7 914 |
- Всего сейчас7 914
COA tackles trading-system parameter tuning where brute-force search fails: a noisy, multimodal profit landscape that punishes single-swarm convergence. Instead of one population collapsing to one point, COA splits candidates into mostly stable packs. Each pack follows its own alpha (local best) and a “culture” defined by the coordinate-wise median, which resists outliers and keeps exploration grounded. Updates combine two pulls: toward the alpha and toward the median, using random pack members as references. Step size shrinks naturally as a pack tightens, and moves are accepted only when fitness improves. Diversity is reinforced by per-pack pup generation: crossover between two parents with a low, dimension-scaled mutation rate. Rare inter-pack swaps spread successful traits without homogenizing the search, making it practical for robust MT5 optimizations. 👉 Read | Forum | @mql5dev
Открыть в TelegramДинамика просмотровСкрыть динамику
Рост просмотров
Часов после публикации
Показать таблицей
| Часов после публикации | Просмотры |
|---|---|
| 0 ч | 0 |
| 1 ч | 2 153 |
| 13 ч | 6 675 |
| 25 ч | 7 340 |
- Через 1 час2 153
- Через 24 часа7 340
- Всего сейчас7 340
Financial tick streams remain difficult to model with fixed-interval sampling. Event-driven architectures fit better when activity is irregular and noise dominates. SDformerFlow combines spiking neurons with windowed attention, using shifted windows to limit attention cost while still capturing broader dependencies. Spikes trigger computation only on significant changes, reducing redundant work. During an MQL5 port, the windowed layout and sparse activation mapped well to OpenCL parallelization. Encoder work centered on SDSA blocks and SPM patch-merge stages. Current work moves to a U-shaped design with symmetric encoder/decoder flow. Instead of separate branches, a nested U-block is used to simplify tensor routing and maintenance. A new CNeuronSDformerUBlock encapsulates encoder, neck recursion, concatenation, and decoder, terminating in an MS-ResBlock... 👉 Read | Freelance | @mql5dev
Открыть в TelegramДинамика просмотровСкрыть динамику
Рост просмотров
Часов после публикации
Показать таблицей
| Часов после публикации | Просмотры |
|---|---|
| 0 ч | 0 |
| 1 ч | 1 895 |
| 13 ч | 5 573 |
| 25 ч | 6 323 |
- Через 1 час1 895
- Через 24 часа6 323
- Всего сейчас6 323
TradingCostPlanner is an MQL5 Expert Advisor that estimates how much of an expected favorable move remains after spread. It uses the chart symbol’s live Bid/Ask plus an exit-spread assumption, producing scenario outputs in points, not profit, and excluding commission, swap, and slippage. Two spread-only estimates are compared. One assumes the current spread at exit. The other uses the greater of the current spread and a percentile computed from recent tick-history samples. Both apply the planned midpoint move, where entry and exit each contribute half of their assumed spreads. The EA validates inputs, checks volume against symbol constraints, rejects stale quotes, and refreshes tick history on a timer. A collapsible panel separates unavailable data from valid projections and shows execution readiness independently. An optional Execute action can submit a mar... 👉 Read | CodeBase | @mql5dev
Открыть в TelegramДинамика просмотровСкрыть динамику
Рост просмотров
Часов после публикации
Показать таблицей
| Часов после публикации | Просмотры |
|---|---|
| 0 ч | 0 |
| 3 ч | 3 772 |
| 15 ч | 6 167 |
| 27 ч | 6 797 |
| 68 ч | 7 949 |
- Через 24 часа6 797
- Всего сейчас7 949
A free MT5 Risk/Reward Calculator script is available for position sizing and risk control. It calculates lot size from a defined risk percentage and stop-loss distance, then renders an on-chart dashboard to avoid manual math. The panel shows computed lot size, monetary risk in account currency, stop loss, take profit, and the resulting risk-to-reward ratio. Calculation can be based on account balance or equity. Instrument handling is automatic by reading tick value, tick size, and broker lot step, with support across FX, metals, indices, and crypto. Inputs cover risk percentage, stop-loss and take-profit distances, balance vs equity selection, and basic display options such as panel position and colors. The script is informational only and does not place orders. 👉 Read | Freelance | @mql5dev
Открыть в TelegramДинамика просмотровСкрыть динамику
Рост просмотров
Часов после публикации
Показать таблицей
| Часов после публикации | Просмотры |
|---|---|
| 0 ч | 0 |
| 3 ч | 3 190 |
| 15 ч | 5 386 |
| 27 ч | 6 078 |
| 68 ч | 7 436 |
- Через 24 часа6 078
- Всего сейчас7 436