Abstract: Digital image segmentation is provided. The method comprises training a neural network for image segmentation with a labeled training dataset from a first domain, wherein a subset of nodes in the neural net are dropped out during training. The neural network receives image data from a second, different domain. A vector of N values that sum to 1 is calculated for each image element, wherein each value represents an image segmentation class. A label is assigned to each image element according to the class with the highest value in the vector. Multiple inferences are performed with active dropout layers for each image element, and an uncertainty value is generated for each image element. Uncertainty is resolved according to expected characteristics. The label of any image element with an uncertainty above a threshold is replaced with a new label corresponding to a segmentation class based on domain knowledge.

Abstract: A method for digital image segmentation is provided. The method comprises training a neural network for image segmentation with a labeled training dataset from a first domain, wherein a subset of nodes in the neural net are dropped out during training. The neural network receives image data from a second, different domain. A vector of N values that sum to 1 is calculated for each image element, wherein each value represents an image segmentation class. A label is assigned to each image element according to the class with the highest value in the vector. Multiple inferences are performed with active dropout layers for each image element, and an uncertainty value is generated for each image element. The label of any image element with an uncertainty value above a predefined threshold is replaced with a new label corresponding to the class with the next highest value.

Abstract: A method for digital image segmentation is provided. The method comprises training a neural network for image segmentation with a labeled training dataset from a first domain, wherein a subset of nodes in the neural net are dropped out during training. The neural network receives image data from a second, different domain. A vector of N values that sum to 1 is calculated for each image element, wherein each value represents an image segmentation class. A label is assigned to each image element according to the class with the highest value in the vector. Multiple inferences are performed with active dropout layers for each image element, and an uncertainty value is generated for each image element. The label of any image element with an uncertainty value above a predefined threshold is replaced with a new label corresponding to the class with the next highest value.

Abstract: A system, method, and computer program product for dynamic, cost effective reallocation of assets among a plurality of investment products comprising a processor, a memory and a computer program stored in the memory. The computer program implementing the present invention controls the reallocation of assets to reduce the transactions costs associated with rebalancing the investor's composite assets according to a composite asset allocation model. Information relating to the composite asset allocation model, the investor's assets, and the investor are stored in memory. Periodically, or upon occurrence of an event, the composite assets are evaluated to determine if rebalancing is necessary. If rebalancing is necessary, the transaction costs associated with the available transactions for performing the rebalancing are compared to select the most economically favorable transaction.

Abstract: A system, method, and computer program product for dynamic, cost effective reallocation of assets among a plurality of investment products comprising a processor, a memory and a computer program stored in the memory. The computer program implementing the present invention controls the reallocation of assets to reduce the transactions costs associated with rebalancing the investor's composite assets according to a composite asset allocation model. Information relating to the composite asset allocation model, the investor's assets, and the investor are stored in memory. Periodically, or upon occurrence of an event, the composite assets are evaluated to determine if rebalancing is necessary. If rebalancing is necessary, the transaction costs associated with the available transactions for performing the rebalancing are compared to select the most economically favorable transaction.

Abstract: A system, method, and computer program product for dynamic, cost effective reallocation of assets among a plurality of investment products comprising a processor, a memory and a computer program stored in the memory. The computer program implementing the present invention controls the reallocation of assets to reduce the transactions costs associated with rebalancing the investor's composite assets according to a composite asset allocation model. Information relating to the composite asset allocation model, the investor's assets, and the investor are stored in memory. Periodically, or upon occurrence of an event, the composite assets are evaluated to determine if rebalancing is necessary. If rebalancing is necessary, the transaction costs associated with the available transactions for performing the rebalancing are compared to select the most economically favorable transaction.