Abstract: Systems, methods, and computer program products for cluster data in a computing network are provided. Different computing devices in a network generate data structures. These data structures have different attributes. A machine learning system is provided to cluster the data structures with the same attributes into clusters. In this way, data structures that include image data are clusters according to the same images. Data structures that include network configuration data are clusters according to the same network configuration data. Data structures that include textual documents are clusters according to the same textual documents or corresponding abstracts.

Abstract: A fund management system permits an investor to transmit a custom creation basket of securities to the fund management system in exchange for shares of a fund. The fund management system receives an available basket of securities from the investor and a fund designated for share creation. The fund management system filters the available securities to determine securities eligible for inclusion in the fund management system. An optimization algorithm determines a quantity of securities that maximizes the utility of adding the securities to the fund, which generates a custom creation basket. The fund management system may receive available baskets from multiple investors and a designation of multiple funds. The fund management system thus may generate a multi-party and multi-fund optimized custom creation basket.