Method and apparatus for managing machine learning process

A method for managing a machine learning process, comprising:

• obtaining, by a processor, a first measurement value representing prediction performance of a model that has been learned by using a first parameter value and training data of first size, the first parameter value being among a set of parameter values applicable to a machine learning algorithm;
• calculating, by the processor and based on the first measurement value, a first expected value and a first variance degree of prediction performance of a model that would be learned by using the first parameter value and training data of second size;
• obtaining, by the processor, a second measurement value representing prediction performance of a model that has been learned by using a second parameter value and training data of the first size, the second parameter value being among the set of parameter values;
• calculating, by the processor and based on the second measurement value, a second expected value and a second variance degree of prediction performance of a model that would be learned by using the second parameter value and training data of the second size;
• and
• calculating, by the processor and based on the first and second expected values and the first and second variance degrees, a third expected value and a third variance degree of prediction performance of a model that would be learned by using a third parameter value and training data of the second size, the third parameter value being among the set of parameter values.

An apparatus that manages a machine learning process, the apparatus comprising:

• a memory configured to store therein a first measurement value and a second measurement value, the first measurement value representing prediction performance of a model that has been learned by using a first parameter value and training data of first size, the first parameter value being among a set of parameter values applicable to a machine learning algorithm, the second measurement value representing prediction performance of a model that has been learned by using a second parameter value and training data of the first size, the second parameter value being among the set of parameter values;
• and
• a processor, coupled to the memory, configured to perform a procedure including:
• calculating, based on the first measurement value, a first expected value and a first variance degree of prediction performance of a model that would be learned by using the first parameter value and training data of second size;
• calculating, based on the second measurement value, a second expected value and a second variance degree of prediction performance of a model that would be learned by using the second parameter value and training data of the second size;
• and
• calculating, based on the first and second expected values and the first and second variance degrees, a third expected value and a third variance degree of prediction performance of a model that would be learned by using a third parameter value and training data of the second size, the third parameter value being among the set of parameter values.