Abstract: A method for an online, task-aware opponent modeling in autonomous racing is described. The method includes concurrently training an opponent-aware policy and an opponent-aware encoder using reinforcement learning. The method also includes calculating, by the opponent-aware encoder, opponent encoding information according to prior opponent positions. The method further includes updating learning parameters of the opponent-aware policy using the opponent encoding information from the opponent-aware encoder to predict actions. The method also includes updating a posterior network according to an auxiliary mutual information loss between the actions predicted by the opponent-aware policy and the opponent encoding information from the opponent-aware encoder.

Abstract: Aspects of the subject disclosure may include, for example, trained, rule based models used to predict whether fiber connections exist upstream of an optical terminal. A regression model is built using optical signal measurements at optical terminals and distances between the optical terminals and an upstream optical terminal. The regression model may be used to predict an optical signal level at an under-test optical terminal based on a distance between the under-test optical terminal and the upstream optical terminal. Other embodiments are disclosed.

Abstract: The system allows real time collection and processing of massive data from many network elements. It has an elastic architecture that scales horizontally to support different network sizes. The system in a uniform data format for downstream consumption. It employs a pub/sub data distribution mechanism that supports multiple concurrent downstream subscribers efficiently in real-time.

Abstract: The system allows real time collection and processing of massive data from many network elements. It has an elastic architecture that scales horizontally to support different network sizes. The system in a uniform data format for downstream consumption. It employs a pub/sub data distribution mechanism that supports multiple concurrent downstream subscribers efficiently in real-time.

Abstract: The system allows real time collection and processing of massive data from many network elements. It has an elastic architecture that scales horizontally to support different network sizes. The system in a uniform data format for downstream consumption. It employs a pub/sub data distribution mechanism that supports multiple concurrent downstream subscribers efficiently in real-time.

Abstract: The system allows real time collection and processing of massive data from many network elements. It has an elastic architecture that scales horizontally to support different network sizes. The system in a uniform data format for downstream consumption. It employs a pub/sub data distribution mechanism that supports multiple concurrent downstream subscribers efficiently in real-time.

Abstract: The system allows real time collection and processing of massive data from many network elements. It has an elastic architecture that scales horizontally to support different network sizes. The system in a uniform data format for downstream consumption. It employs a pub/sub data distribution mechanism that supports multiple concurrent downstream subscribers efficiently in real-time.

Abstract: The system allows real time collection and processing of massive data from many network elements. It has an elastic architecture that scales horizontally to support different network sizes. The system in a uniform data format for downstream consumption. It employs a pub/sub data distribution mechanism that supports multiple concurrent downstream subscribers efficiently in real-time.

Abstract: The system allows real time collection and processing of massive data from many network elements. It has an elastic architecture that scales horizontally to support different network sizes. The system in a uniform data format for downstream consumption. It employs a pub/sub data distribution mechanism that supports multiple concurrent downstream subscribers efficiently in real-time.

Abstract: A method for synthesis of porous inorganic materials, preparation of a catalyst and catalytic cracking of petroleum hydrocarbons thereof includes processes for synthesis of porous inorganic materials and preparation of the catalytic cracking catalyst and catalytic cracking of petroleum hydrocarbons. The synthesis process is advantaged in low cost in raw materials; the porous inorganic material has various pore structures; and transitional metal used overcomes the defects of the catalytic properties. The porous inorganic material serving as the main active ingredient and containing crystalline aluminum silicate zeolite structures provides surface acidity required by the catalytic reaction. The surface acidity is flexibly adjusted. The hierarchical pore profile improves the accessibility of the active center of the zeolite structure and favors the reaction efficiency and benefits of the petroleum hydrocarbon cracking, and reduces the negative effects caused by diffusion limit.

Abstract: A property manager layer retrieves from a persistence layer values of properties used by a software application and provides the values to the application at run time. A user interface layer accepts user input to update at least one of the properties. The persistence layer is responsive to the user interface layer to update the properties based on the user input. The user interface layer notifies the property manager layer that the at least one of the properties has been updated. The property manager layer notifies a property listener for the software application if any updated property value is used by the software application. The property listener modifies at least one program variable in the software application that is dependent on an updated property value without restarting the software application.