Matrix multiplication:
tf.matmul(x,y) matrix multiplication: the shape of A is (2,3), the shape of B is (3,4), and the result shape is (2,4)
Operations on matrices and matrix elements:
Add matrix elements:
tf.math.add(x,y) is equivalent to x + y
Matrix subtraction:
tf.math.subtract(x,y) is equivalent to xy
Multiply matrix elements:
tf.math.multiply(x,y) is equivalent to: x*y
Matrix elementwise division:
tf.math.devide(x,y) is equivalent to x/y
Absolute value of matrix elements:
tf.math.abs(x)
Comparison operation:
tf.math.maximum(a,b)
tf.math.minimum(a,b)
tf.math.equal(a,b)
tf.math.equal(a,3)
tf.math.equal(3,3)
tf.math.equal(a,a)
tf.math.greater(a,3)
tf.math.greater_equal(a,3)
Operations on matrices and numbers:
Multiplication of matrix elements: 3*a is equivalent to a*3
Adding the number of matrix elements: 3+a is equivalent to a+3
Subtract the number of matrix elements: 3-a is equivalent to a-3
Divide the number of matrix elements by a/3,3/a
Similar computational operations for matrix elements are in the tf.math package
special:
Maximum value:
tf.math.reduce_max(a,axis=1,keepdims=True) will be the maximum value of the last dimension, you can choose to keep the dimension or compress it
The position indices corresponding to the maximum value:
tf.argmax(predict_result["logits"],axis=-1) Compress the last dimension to find the position corresponding to the maximum value
Find the maximum value of top_k:
tf.math.top_k(x)
tf.where operation:
Set the vector greater than 0.5 to 1, and less than 0.5 to 0
tf.where(A>0.5, x=1, y=0) indicates that the element in A>0.5 whose condition is true is set to x, and the element which is false is set to y