Fair deep reinforcement learning is provided. A microstate of an environment and reaction of items in a plurality of microstates within the environment are observed after an agent performs an action in the environment. Semi-supervised training is utilized to determine bias weights corresponding to the action for the microstate of the environment and the reaction of the items in the plurality of microstates within the environment. The bias weights from the semi-supervised training are merged with non-bias weights using an artificial neural network. Over time, it is determined where bias is occurring in the semi-supervised training based on merging the bias weights with the non-bias weights in the artificial neural network. A deep reinforcement learning model that decreases reliance on the bias weights is generated based on determined bias to increase fairness.

According to an embodiment, disclosed is a system comprising a processor wherein the processor is configured to receive an input data comprising an image of an ocular region of a user, clinical data of the user, and external factors; extract, using an image processing module comprising adaptive filtering techniques, ocular characteristics, combine, using a multimodal fusion module, the input data to determine a holistic health embedding; detect, based on a machine learning model and the holistic health embedding, a first output comprising likelihood of myopia, and severity of myopia; predict, based on the machine learning model and the holistic health embedding, a second output comprising an onset of myopia and a progression of myopia in the user; and wherein the machine learning model is a pre-trained model; and wherein the system is configured for myopia prognosis powered by multimodal data.

Embodiments discussed herein facilitate determination of risk of relapse of AML post-transplant. One example embodiment is a method, comprising: accessing a digital whole slide image (WSI) comprising a post-transplant bone marrow aspirate from a patient that has acute myeloid leukemia (AML); segmenting one or more myeloblasts on the digital WSI; extracting one or more features from the segmented one or more myeloblasts; providing the one or more features extracted from the segmented one or more myeloblasts to a trained machine learning model; and receiving, from the trained machine learning model, an indication of a risk of relapse of the AML.

In an approach for building a machine learning model that predicts the appropriate action to resolve a malfunction or system error, a processor receives an alert that a malfunction or a system error has occurred. A processor creates a workspace on a ChatOps platform integrated with a chatbot and one or more tools. A processor inputs data relating to the alert in a natural language format. A processor processes the data using a natural language processing algorithm. Responsive to determining a pre-set threshold for outputting the appropriate action is not met, a processor establishes a conversation between two or more support service agents in the workspace. A processor monitors the conversation using the natural language processing algorithm. A processor analyzes a transcript of the conversation using text analytics or pattern matching. A processor creates and trains a machine learning model to predict the appropriate action in future iterations.