Abstract: Computer-implemented techniques for learning composite machine learned models are disclosed. Benefits to implementors of the disclosed techniques include allowing non-machine learning experts to use the techniques for learning a composite machine learned model based on a learning dataset, reducing or eliminating the explorative trial and error process of manually tuning architectural parameters and hyperparameters, and reducing the computing resource requirements and model learning time for learning composite machine learned models. The techniques improve the operation of distributed learning computing systems by reducing or eliminating straggler effects and by reducing or minimizing synchronization latency when executing a composite model search algorithm for learning a composite machine learned model.

Abstract: Systems and methods determine optimized hyperparameter values for one or more machine-learning models. A sample training data set from a larger corpus of training data is obtained. Initial hyperparameter values are then randomly selected. Using the sample training data set and the randomly chosen hyperparameter values, an initial set of performance metric values are obtained. Maximized hyperparameter values are then determined from the initial set of hyperparameter values based on the corresponding performance metric value. A larger corpus of training data is then evaluated using the maximized hyperparameter values and the corresponding machine-learning model, which yields another corresponding set of performance metric values. The maximized hyperparameter values and their corresponding set of performance metric values are then merged with the prior set of hyperparameter values.

Abstract: Embodiments of the disclosed technologies provide tree-based transfer learning of hyperparameters of a machine learning model or tunable parameters of a black box system. A similar reference task tree is selected from a set of reference task trees. Data is transferred from the similar reference task tree to a target task tree.

Abstract: Computer-implemented techniques for learning composite machine learned models are disclosed. Benefits to implementors of the disclosed techniques include allowing non-machine learning experts to use the techniques for learning a composite machine learned model based on a learning dataset, reducing or eliminating the explorative trial and error process of manually tuning architectural parameters and hyperparameters, and reducing the computing resource requirements and model learning time for learning composite machine learned models. The techniques improve the operation of distributed learning computing systems by reducing or eliminating straggler effects and by reducing or minimizing synchronization latency when executing a composite model search algorithm for learning a composite machine learned model.

Abstract: In an example embodiment, training data is obtained, the training data comprising values for a plurality of different features. Then a global machine learned model is trained using a first machine learning algorithm by feeding the training data into the first machine learning algorithm during a fixed effect training process. A non-linear first random effects machine learned model is trained by feeding a subset of the training data into a second machine learning algorithm, the subset of the training data being limited to training data corresponding to a particular value of one of the plurality of different features.

Abstract: Systems and methods determine optimized hyperparameter values for one or more machine-learning models. A sample training data set from a larger corpus of training data is obtained. Initial hyperparameter values are then randomly selected. Using the sample training data set and the randomly chosen hyperparameter values, an initial set of performance metric values are obtained. Maximized hyperparameter values are then determined from the initial set of hyperparameter values based on the corresponding performance metric value. A larger corpus of training data is then evaluated using the maximized hyperparameter values and the corresponding machine-learning model, which yields another corresponding set of performance metric values. The maximized hyperparameter values and their corresponding set of performance metric values are then merged with the prior set of hyperparameter values.

Abstract: A two-terminal photon-effect transistor (PET) is described that simplifies the photo sensing pixel by combing photodiode and field effect transistor dual functions into one simple but effective unit. Photons excite electrons from the valance band of semiconducting material as the electrode-free gate to modulate resistivity between source and drain, which directly results in current amplification of photo signal without traditional photo-electrical conversion and electrical amplification dual processes. PET possesses significance in both structural simplification and functional enhancement. As an implementing example of PET, a nanowire camera (NC) with large sensing area and extremely high resolution is fabricated by integrating millions of vertically aligned nanowire arrays in-between of orthogonal top and bottom nano-stripe electrodes. Each nanowire works as independent three-dimensional (3D) PET pixel, enabling the NC an ultra-high resolution and much simplified architecture.

Abstract: A two-terminal photon-effect transistor (PET) is described that simplifies the photo sensing pixel by combing photodiode and field effect transistor dual functions into one simple but effective unit. Photons excite electrons from the valance band of semiconducting material as the electrode-free gate to modulate resistivity between source and drain, which directly results in current amplification of photo signal without traditional photo-electrical conversion and electrical amplification dual processes. PET possesses significance in both structural simplification and functional enhancement. As an implementing example of PET, a nanowire camera (NC) with large sensing area and extremely high resolution is fabricated by integrating millions of vertically aligned nanowire arrays in-between of orthogonal top and bottom nano-stripe electrodes. Each nanowire works as independent three-dimensional (3D) PET pixel, enabling the NC an ultra-high resolution and much simplified architecture.

Abstract: Various examples are provided for pillar array photo detectors. In one example, among others, a photo detection system includes an array of substantially aligned photo sensitive nanorods extending between first and second electrodes, and a plurality of resistance monitoring circuits coupled at different positions about the circumference of the electrodes. In another example, a photo detector includes first and second electrodes, and an array of substantially aligned photo sensitive nanorods extending between the substantially parallel electrodes. Light passing through an electrode excites electrons in the photo sensitive nanorods that are illuminated by the light.

Abstract: Systems and methods are disclosed that describe flexible photo-sensing pixels and interconnects that have comparable pixel densities and functionality as the human retina. The pixels comprise vertically aligned, nanowire cluster piles that serve as the three-dimensionally compressible photoreceptor pixels, and flexible, transparent interconnected electrodes. Shape-adaptive high-resolution optic-electrical imaging system are described that can serve as a human retina and a retinal prosthesis for restoring vision.

Abstract: A two-terminal photon-effect transistor (PET) is described that simplifies the photo sensing pixel by combing photodiode and field effect transistor dual functions into one simple but effective unit. Photons excite electrons from the valance band of semiconducting material as the electrode-free gate to modulate resistivity between source and drain, which directly results in current amplification of photo signal without traditional photo-electrical conversion and electrical amplification dual processes. PET possesses significance in both structural simplification and functional enhancement. As an implementing example of PET, a nanowire camera (NC) with large sensing area and extremely high resolution is fabricated by integrating millions of vertically aligned nanowire arrays in-between of orthogonal top and bottom nano-stripe electrodes. Each nanowire works as independent three-dimensional (3D) PET pixel, enabling the NC an ultra-high resolution and much simplified architecture.

Abstract: Various examples are provided for pillar array photo detectors. In one example, among others, a photo detection system includes an array of substantially aligned photo sensitive nanorods extending between first and second electrodes, and a plurality of resistance monitoring circuits coupled at different positions about the circumference of the electrodes. In another example, a photo detector includes first and second electrodes, and an array of substantially aligned photo sensitive nanorods extending between the substantially parallel electrodes. Light passing through an electrode excites electrons in the photo sensitive nanorods that are illuminated by the light.