Method and data processing device for ascertaining properties of hair colors in a computer-assisted manner

A method for computer-aided determination of properties of hair color, including:

• preparing hair color data, wherein the hair color data has values for a multitude of hair pre-condition parameters and for at least one coloring result parameter for a multitude of coloring processes;
• wherein the multitude of color pre-condition parameters has a first concentration of a first dye precursor and a second concentration of a second dye precursor for each coloring process of a multitude of coloring processes;
• wherein a measured specification about a property of hair color is provided for each coloring process of the multitude of coloring processes of the at least one coloring result parameter including a brightness of the hair color, resulting from the pre-condition parameters;
• determining, via a processor, a relationship between the multitude of color pre-condition parameters and the at least one coloring result parameter by employing predictive analytics based on the hair color data, using a tree-ensemble learner, implemented by the processor, that includes a plurality of decision tree ensembles;
• and
• providing a display of results pertaining to the relationship via a display device, including the brightness of the hair color, resulting from the pre-condition parameters.

A method for computer-aided determination of properties of hair color, including:

• preparing hair color data, wherein the hair color data has values for a plurality of hair pre-condition parameters and for at least one coloring result parameter for a plurality of coloring processes, wherein the at least one coloring result parameter includes a brightness of the hair color;
• wherein the plurality of color pre-condition parameters has a first concentration of a first dye precursor and a second concentration of a second dye precursor for each coloring process of a plurality of coloring processes;
• wherein a measured specification about a property of hair color is provided for each coloring process of the plurality of coloring processes of the at least one coloring result parameter,
• loading the hair color data to a non-transitory computer readable storage medium, with each of parameter of the hair color data loaded into a dedicated column;
• generating, via the processor, a column filter for selecting certain columns of hair color data as color pre-conditions parameters and coloring result parameters, generating training data;
• determining, via a processor, a relationship between the plurality of color pre-condition parameters and the at least one coloring result parameter, including the brightness for the hair color, by employing predictive analytics based on the hair color data, using a tree-ensemble learner, implemented by the processor, that includes a plurality of decision tree ensembles, wherein the step of generating the relationship comprises generating, via the processor, the relationship via application of the tree-ensemble learner to the selected columns of pre-conditions parameters and coloring result parameters of the training data;
• applying the tree-ensemble learner, via the processor, in conjunction with the generated relationship based on the training data, to remaining data of the hair color that was not part of the selected columns for the training data, to thereby predict one or more of the color result parameters including the brightness for the hair color, corresponding to the color pre-conditions parameters for the remaining data;
• and
• providing the one or more of the predicted color result parameters, including the brightness for the hair color, on a display via a display device.