Abstract: A fractal power converter and a method for constructing the fractal power converter, the method for constructing the fractal converter including: replacing a switch and/or an energy storage element of an asymmetric half-bridge sub-module with the same asymmetric half-bridge sub-module for at least two times, replacing a switch and/or an energy storage element of a symmetric half-bridge sub-module with the same symmetric half-bridge sub-module for at least two times, and replacing a switch and/or an energy storage element of an H-bridge sub-module with the same H-bridge sub-module for at least two times. Compared with a conventional high-voltage and large-current power converter, the fractal power converter may output any high-voltage and large-current waveform. The system is modularized, the structure is simple and extensible, and voltage and current may be evenly distributed among the modules. The control method is simple and easy to standardize. Multi-port parallel output is allowed.

Abstract: A deep relational factorization machine (“DRFM”) system is configured to provide a high-order prediction based on high-order feature interaction data for a dataset of sample nodes. The DRFM system can be configured with improved factorization machine (“FM”) techniques for determining high-order feature interaction data describing interactions among three or more features. The DRFM system can be configured with improved graph convolutional neural network (“GCN”) techniques for determining sample interaction data describing sample interactions among sample nodes, including sample interaction data that is based on the high-order feature interaction data. The DRFM system generates a high-order prediction based on the high-order feature interaction embedding vector and the sample interaction embedding vector. The high-order prediction can be provided to a prediction computing system configured to perform operations based on the high-order prediction.