• 概述
  • 评论 (1)
  • 评论 (13)

Neutron star

Overview and Logic

This is my latest Expert Advisor based on neural networks.

One of the major differences is the successful implementation of the martingale technique.

The Expert Advisor uses patterns. The most frequent patterns are used for trading, they help to determine the price movement in the next time moment.


Features

  1. The EA only works with its own orders. The magic numbers of the EA's orders are changed automatically in the range from 123456 to 654321.
  2. After a losing trade the EA increases Lots Martingale times. You can limit the maximum Lots, by changing the Max_Lots input. Once the order volume reaches the Max_Lots value, the EA will start to open orders with a value of Lots.
  3. The EA uses virtual Stop_Loss and Take_Profit. You can also additionally use normal Stop_Loss and Take_Profit.
  4. You can optionally use VPS. Delay does not affect EA's operation.
  5. The EA is very easy to use. It only needs a minimum initial deposit.


Important!!! Preparing for trade

The EA needs training. Necessary stages:

  1. Check Optimization.
  2. Select Model: Open prices only, Timeframes: M1.
  3. Configuring input parameters. Expert properties: check test. Set start=1, step=1, stop=10.
  4. Start optimization. Click Start.


Recommendations

  1. The training process can be applied to any symbol. The EA uses Global Variables. Before training on a new symbol press F3 and delete current global variables.
  2. It is recommended to re-train the EA every 2-3 months.
  3. Max_Lots should be 8-12 times larger than Lots.
  4. It is not recommended to test the EA and trade in real time simultaneously.


Inputs

  1. Lots - the volume of market orders. Recommended value Lots=0.01;
  2. Max_Lots - maximum volume of market orders. Recommended value Max_Lots =0.16;
  3. Martingale - lot increase. Recommended value Martingale =2;
  4. Take Profit. Recommended Take Profit=0.01;
  5. Stop Loss. Recommended Stop Loss =0.01;
  6. test - this parameter is required for the EA training. Recommended value test=1;

Below are training results from 2010 to 2016.

Best regards, Dmitriy.

asaens15
1053
2017.03.01 22:25 
 

Losses incurred after using for a few weeks ... discontinued using it ... too risky for me. I had high hopes for it in the beginning but eventually I had a large loss and didn't want a repeat or know how to resolve it other than to look for safer alternatives.