Python: Advanced Guide to Artificial Intelligence

Defining the loss function

Next, we define the loss function using the mean squared error (MSE). MSE is defined as follows:

More details about MSE can be found from the following links: https://en.wikipedia.org/wiki/Mean_squared_errorhttp://www.statisticshowto.com/mean-squared-error/

The difference in the actual and estimated value of y is known as residual. The loss function calculates the mean of squared residuals. We define it in TensorFlow in the following way:

loss = tf.reduce_mean(tf.square(model - y_tensor))

• model - y_tensor calculates the residuals
• tf.square(model - y_tensor) calculates the squares of each residual
• tf.reduce_mean( ...