Boundary search test support device and boundary search test support method

A boundary search test support method of a boundary search test support device that comprises an information processing device including a storage device configured to store a plurality of predetermined input vectors, the method comprising the steps of:

• acquiring the plurality of input vectors from the storage device, executing a test by sequentially inputting the input vectors to a program generated by a neural network, and acquiring output vectors which are test results corresponding to the input vectors,
• generating a coordinate system having a respective element, of a predetermined plurality of the elements among the elements constituting the output vectors, as each coordinate axis of the coordinate system,
• respectively generating, in the coordinate system, a straight line in which the plurality of elements has a same value and a hyperplane in which a sum of values of the plurality of elements is taken as a predetermined value, and
• arranging in the coordinate system a most antagonistic point which is an intersection point of the straight line and the hyperplane and boundary vectors whose values of the plurality of elements rank higher than or equal to a predetermined ranking among the output vectors, and displaying the coordinate system together with input vectors corresponding to the boundary vectors on a predetermined output device,
• wherein when a number of the plurality of the elements is equal to or less than 3, the method further comprising the steps of:
• displaying the coordinate system on the output device, in which the straight line, the hyperplane, the most antagonistic point, the boundary vectors, and the input vectors corresponding to the boundary vectors are respectively arranged, and
• receiving, by the input device, a displayed number of the boundary vectors, a first coefficient related to distances from the boundary vectors to corresponding nearest points in the straight line, and a second coefficient related to distances from the nearest points of the boundary vectors to the most antagonistic point, adding values obtained by multiplying the distances between the boundary vectors and the nearest points by the first coefficient and the values obtained by multiplying the distances between the nearest points of the boundary vectors and the most antagonistic point by the second coefficient together, and arranging boundary vectors corresponding to the displayed number in the coordinate system in an ascending order of the sums.