Method for designing machine learning model using a 3D user interface and system using same

A method for designing a machine learning model using a 3D user interface, the method comprising:

• a layer input step of inputting a type of a layer and parameters of the layer using a layer input panel including a layer type determination panel and a layer parameter input panel;
• a layer model configuration step of receiving the type and parameters of the layer selected in the layer input step, constructing a layer model using the input layers and displaying a configuration of the layer model on a layer state display panel of a layer state panel;
• a layer model process step of parsing the layer model displayed on the layer state display panel of the layer state panel to calculate a layer position and node position for visualization;
• and
• a layer model 3D visualization step of visualizing the layer model calculated in the layer model process step in 3D, wherein in the layer model 3D visualization step, when defining the shape of the layer as (ny, nz, ch), the layer included in the layer model is represented as ch square units (L) having a height of the ny value and a width of the nz value, and wherein in the layer model 3D visualization step, when the layer model includes two or more layers, the preceding layer is called a previous layer, and the subsequent layer is called a present layer, and a relationship line connecting a point corresponding to one current node for the square unit (L) corresponding to the present layer and a point of at least one previous node belonging to the square unit (L) corresponding to the previous layer required to form the one current node is visualized.