Trixter EE
- Experts
-
Andrei Samokhin
Welcome to my seller page. I develop applications in the fields of AI and machine learning, and I am also passionate about trading systems. My products are based on various machine learning techniques. I research and apply state-of-the-art algorithms for time series analysis and am constantly - Version: 1.0
- Activations: 15
Elliptic Envelope is a parametric machine learning algorithm designed to detect outliers and anomalies in multidimensional data.
The method assumes that "normal" data (the main sample) follows a multidimensional normal (Gaussian) distribution.
Operating Principle
- Ellipsoid Formation: In the feature space, the algorithm builds a hyperellipsoid around the central group of points. The boundary of this ellipsoid is determined by the covariance matrix and the mean values of the features.
- Use of Mahalanobis Distance: For each point, the Mahalanobis distance to the center of the distribution is calculated. This distance takes into account not only proximity, but also the correlation between features and the variance of the data along different axes.
- Robustness Estimation (MCD): To prevent outliers from distorting the estimation of the ellipsoid's center and shape, the method uses the Minimum Covariance Determinant (MCD). The algorithm searches for a subset of data of a specific size (specified by a parameter) that minimizes the determinant of the covariance matrix, excluding extreme values from the ellipsoid fitting process.
- Classification:
- Points inside the ellipsoid are normal objects (inliers, label 1).
- Points outside the ellipsoid are anomalies or outliers (outliers, label -1).
- The distance threshold beyond which a point is considered an outlier is defined by the contamination hyperparameter.
What It Is Used For
- Data Cleaning: Removing extreme noise and measurement errors prior to training primary models (classification, regression).
- Anomaly / Outlier Detection: Finding non-standard objects in a dataset, for example:
- Fraud Detection.
- Filtering sensor or equipment failures.
- Filtering atypical patterns or noise in features.
The Trixter EE bot is built using this approach.
The Trixter EE algorithm implements pattern recognition using the Elliptic Envelope (EE for short). We operate on the assumption that the gold market is fairly efficient and that most price fluctuations form market noise that cannot be predicted. However, we can isolate fragments of the gold chart (so-called anomalies) where it is highly predictable. In this way, I have identified 40% of chart areas that are well predictable.
Key Principles
- Analysis of gold volatility across multiple timeframes and indicator periods.
- Volatility grouping to search for effective patterns (ranging from 100 to 3000).
- Statistical verification: high expected payoff and statistical significance.
- Patterns that fail verification are marked as noise; no trades are opened in such situations.
- Mandatory 2-year forward test prior to deployment on a real account.
Trading Logic Features
- Entry with limit orders at a chosen distance from the price, ensuring minimal slippage.
- If a limit order is not triggered (price did not touch the level) and the signal vanishes, the order is deleted.
- Open positions are accompanied by Stop Loss and Take Profit levels and are closed at market price upon a counter signal.
- The robot does not use martingale, arbitrage, or other high-risk strategies.
- Recommended timeframes: from M5 to H1. Optimal: H1 XAUUSD.
- Flexible trading activity configuration is provided (parameters below).
Main Settings
- Allow BUY signals, Allow SELL signals — allow or disable buy/sell trades when confident in the market direction.
- Signals sensitivity, Filter sensitivity — threshold for signal sensitivity and noise filtering. Default value is 0.5. Can be increased to 0.7–0.9 for cleaner signals.
- Filter by trading hours — restriction on trading time. Default is around the clock.
Capital Management Settings
- Distance in points for limit order — distance from the price to the limit order. Default: 250. Lower value → more trades; higher value → more conservative.
- Max positions + orders number — maximum number of simultaneously open positions (no grid/averaging). Default: 3.
- Additionally configurable: delay between openings, maximum spread control, Stop Loss, and Take Profit levels.
Important: Default settings are already applied. Do not change them unless necessary. Save the standard set file to be able to revert to factory parameters.
Elliptic envelope + two classifiers → export to ONNX | Forward period: 2 years | Rec. timeframe: H1 XAUUSD
