AI talking

Quoting joyny

Disliked

{quote} Here how AI drawdowns look.. 80 days sit in DD.. that's real trading

Ignored

Quoting joyny

Disliked

Won't use python AIs.. no stable results found. I will use the simple neurons which is possible to implement directly in the MQL5 expert advisor and use optimizations to find neuron weights.

Ignored

Quoting ryuryu

Disliked

{quote} That's right, it has an eternity ahead

Ignored

Quoting dantpm

Disliked

{quote} Did you consider trying the TensorFlow stock price prediction algorithm? It is on the TensorFlow website and is able to predict stock prices several months in advance. My thought was that it would be pretty simple to amend the code for Forex and shorter timeframes. It does use Python with the TensorFlow software so that might be an issue.

Ignored

Quoting joyny

Disliked

{quote} I did myself neural network in python with TensorFlow. But spent not much time therefore maybe can't expect results. Maybe you talk about this TensorFlow python solution for stock price prediction? I implemented it,

Ignored

Quote

Disliked

The number of epochs required to train a TensorFlow model depends on various factors, such as the complexity of the model, the size of the dataset, and the computational resources available. In general, it is recommended to train the model until the loss function converges and does not improve significantly with additional epochs.

Quote

Disliked

Seven epochs may be sufficient to train some models, but it may not be enough for others. It is important to monitor the loss function during training and validate the model's performance on a separate validation dataset. If the model achieves good performance and the loss function has converged after seven epochs, then you may use the results of the model produced with seven epochs. However, it is always a good practice to experiment with different numbers of epochs and evaluate the model's performance to ensure optimal results.

Quote

Disliked

If you have observed that the model's performance has stabilized after a certain number of epochs and the prediction accuracy does not change significantly (within +/- 1-2%), then it can be a good indication that the model is sufficiently trained and further training is not necessary. However, it is still a good practice to perform additional validation tests on a separate test dataset to ensure that the model's performance is consistent and reliable.

Quoting joyny

Disliked

did some EA which trades with 10 neurons, input data is RSI indicator, fixed SL/TP.

Ignored

Quoting pizd0s

Disliked

While neural networks can be a powerful tool for predicting market prices, they are not always the best choice for forex trading. One reason is that neural networks can be prone to overfitting, which means they may fit too closely to historical data and perform poorly when used to predict future prices.

Ignored

Quoting pizd0s

Disliked

Additionally, neural networks can be complex and difficult to interpret, which can make it challenging for traders to understand how the model arrived at its predictions. This lack of transparency can be problematic, especially if traders are relying solely on the model's output to make trading decisions.

Ignored

// EntryConditionMet
bool EntryConditionMet(ENUM_ORDER_TYPE &orderType)
{
  if (!IsTimeInRange(time_from_h, time_to_h))
  {
    Print("Trading not allowed outside of specific time span");
    return false;
  }
  
  MqlRates rates[];
  ArraySetAsSeries(rates, true);
  if (CopyRates(_Symbol, TimeFrame, 1, 4, rates) == -1)
  {
    Print("Error copying rates: ", GetLastError());
    return false;
  }
  // Check for bullish Three Bar Reversal pattern
  if (rates[2].low < rates[1].low && rates[2].low < rates[3].low && rates[0].close > rates[0].open)
  {
    orderType = ORDER_TYPE_BUY;
    return true;
  }
  // Check for bearish Three Bar Reversal pattern
  else if (rates[2].high > rates[1].high && rates[2].high > rates[3].high && rates[0].close < rates[0].open)
  {
    orderType = ORDER_TYPE_SELL;
    return true;
  }
  return false;
}

Quote

Disliked

This entry condition checks for a Three Bar Reversal pattern. If the pattern is bullish (middle bar has the lowest low and the last bar is bullish), the expert advisor will enter a long position (buy). If the pattern is bearish (middle bar has the highest high and the last bar is bearish), the expert advisor will enter a short position (sell).

// EntryConditionMet
bool EntryConditionMet(ENUM_ORDER_TYPE &orderType)
{
  if (!IsTimeInRange(time_from_h, time_to_h))
  {
    Print("Trading not allowed outside of specific time span");
    return false;
  }
  
  MqlRates rates[];
  ArraySetAsSeries(rates, true);
  if (CopyRates(_Symbol, TimeFrame, 1, 3, rates) == -1)
  {
    Print("Error copying rates: ", GetLastError());
    return false;
  }
  // Check for bullish Inside Bar pattern
  if (rates[0].high < rates[1].high && rates[0].low > rates[1].low && rates[1].close > rates[1].open)
  {
    orderType = ORDER_TYPE_BUY;
    return true;
  }
  // Check for bearish Inside Bar pattern
  else if (rates[0].high < rates[1].high && rates[0].low > rates[1].low && rates[1].close < rates[1].open)
  {
    orderType = ORDER_TYPE_SELL;
    return true;
  }
  return false;
}

Quote

Disliked

This entry condition checks for an Inside Bar pattern. If the Inside Bar pattern is bullish (current bar's close is higher than its open), the expert advisor will enter a long position (buy). If the Inside Bar pattern is bearish (current bar's close is lower than its open), the expert advisor will enter a short position (sell).

Quote

Disliked

Prepare something compleatly out of the box which is not known from theory and not found in internet - be creative

Quote

Disliked

I present to you a creative and unique entry logic based on the "Curved Bar" pattern. This pattern considers a combination of three consecutive bars that form a curve-like structure. The pattern is formed when the middle bar is either higher or lower than both the previous and next bar, and the price difference between the first and the third bar is minimal.

// Curvature threshold to be adjusted for optimization
input double CurvatureThreshold = 0.01;
// EntryConditionMet
bool EntryConditionMet(ENUM_ORDER_TYPE &orderType)
{
  if (!IsTimeInRange(time_from_h, time_to_h))
  {
    Print("Trading not allowed outside of specific time span");
    return false;
  }
  
  MqlRates rates[];
  ArraySetAsSeries(rates, true);
  if (CopyRates(_Symbol, TimeFrame, 1, 4, rates) == -1)
  {
    Print("Error copying rates: ", GetLastError());
    return false;
  }
  double priceDifference = MathAbs(rates[0].close - rates[3].close);
  double middleBarRange = rates[2].high - rates[2].low;
  // Check for bullish Curved Bar pattern
  if (rates[2].low < rates[1].low && rates[2].low < rates[3].low &&
      priceDifference <= middleBarRange * CurvatureThreshold &&
      rates[0].close > rates[0].open)
  {
    orderType = ORDER_TYPE_BUY;
    return true;
  }
  // Check for bearish Curved Bar pattern
  else if (rates[2].high > rates[1].high && rates[2].high > rates[3].high &&
           priceDifference <= middleBarRange * CurvatureThreshold &&
           rates[0].close < rates[0].open)
  {
    orderType = ORDER_TYPE_SELL;
    return true;
  }
  return false;
}

Quote

Disliked

The entry condition checks for a "Curved Bar" pattern. If the pattern is bullish (middle bar has the lowest low and the price difference between the first and the third bar is minimal), the expert advisor will enter a long position (buy). If the pattern is bearish (middle bar has the highest high and the price difference between the first and the third bar is minimal), the expert advisor will enter a short position (sell).

Quoting joyny

Disliked

And you can force AI to be more creative

Ignored

Quoting joyny

Disliked

{quote} Here results for EURUSD for AI proposed strategy "Curved Bar": {image} Here AUDUSD: {image}

Ignored

Quoting joyny

Disliked

Here results for EURUSD for AI proposed strategy "Curved Bar"

Ignored

Quote

Disliked

There are several popular forex patterns based on candles that traders often use to identify potential trading opportunities. Here are some of them:

Quote

Disliked

Doji: This is a candlestick pattern that indicates indecision in the market. It has a small body with upper and lower wicks of equal length, indicating that buyers and sellers are in a state of equilibrium.

Quote

Disliked

Hammer: A hammer is a bullish reversal pattern that has a long lower wick and a small body at the top of the candle. It suggests that sellers are losing control, and buyers are beginning to take over.

Quote

Disliked

Shooting star: A shooting star is a bearish reversal pattern that has a long upper wick and a small body at the bottom of the candle. It suggests that buyers are losing control, and sellers are beginning to take over.

Quote

Disliked

Engulfing pattern: This is a reversal pattern that occurs when a small candle is followed by a larger one that completely engulfs it. A bullish engulfing pattern occurs at the bottom of a downtrend, while a bearish engulfing pattern occurs at the top of an uptrend.

Quote

Disliked

Harami pattern: The harami pattern is a reversal pattern that consists of two candles. The first candle is large and the second is small, and the body of the second candle is completely inside the body of the first. A bullish harami occurs at the bottom of a downtrend, while a bearish harami occurs at the top of an uptrend.