LossLess Neuron

The expert receives signals from an artificial neural network that uses the RSI. Transactions closes the reverse signal. The presence of the signal is checked to close the bar. There is also a function of:

• the change in volume in proportion to the balance of the position change (for the function tester);


• transfer of losing trades to breakeven;

Options

• Start with lot - the initial volume of the position is increased in proportion to the balance of the change;


• Lotsize by balance - Lot increase proportionally change the balance, yes / no. Use only in the tester;


• Timeframe - period of the schedule, the presence of the signal is checked for closure bars of said period;


• a1 ... a8 - input parameters of the neural network, the value from 0 to 100;


• a9 - signal strength of the neural network, in% (range from 20 to 80). The higher the value, the smaller transactions;


• a10 - averaging period RSI indicator;


• Close losses - close unprofitable deals, yes / no. If "no", then when a take profit return signal previous losing trade will be moved to the opening price of the transaction - to breakeven, and if the previous transaction is profitable, it will be closed;


• ...maximum deals with losses - the maximum number of outstanding transactions unprofitable. Transactions will be opened in inverse signals as long as this limit is reached;


• Slippage (in pips) - slip allowable value when the opening of the transaction;


• Magic number - magic number of transactions.

Training and testing

Training the neural network (parameter selection a1 ... a9) Held in EURUSD in the range of from 1 September to 1 November 2016, all the ticks, the M30, the closure of loss-making transactions.

Testing was conducted in the same range, but the transfer loss trades to breakeven (see. Figure).