如何获得该神经元网络的Cuadratic-Error并将其绘制出来？

Question

我需要获取每个步骤（number_of_training_iterations）的明显误差，然后使用例如pandas库对其进行绘制。

我用于训练神经元网络的代码是：

def train(self, training_set_inputs, training_set_outputs, number_of_training_iterations):
        for iteration in range(0, number_of_training_iterations):
            # Neuronal network predicts 
            output_from_hidden_layer, output_from_output_layer = self.think(training_set_inputs)

            # Calculate output layer error
            output_layer_error = training_set_outputs - output_from_output_layer
            output_layer_delta = output_layer_error * self.__sigmoid_derivative(output_from_output_layer)

            # Calculate hidden layer error
            hidden_layer_error = output_layer_delta.dot(self.output_layer.synaptic_weights.T)
            hidden_layer_delta = hidden_layer_error * self.__sigmoid_derivative(output_from_hidden_layer)

            # How much to adjust weights
            hidden_layer_adjustment = training_set_inputs.T.dot(hidden_layer_delta)
            output_layer_adjustment = output_from_hidden_layer.T.dot(output_layer_delta)

            # Adjust weight of each layer
            self.hidden_layer.synaptic_weights += hidden_layer_adjustment
            self.output_layer.synaptic_weights += output_layer_adjustment

下面的代码训练神经网络。但是我不知道如何获得每个步骤的误差并将其绘制在图形中。

任何人都知道我可以采取哪些步骤来实现这一目标？

谢谢。