Abstract: Disclosed are techniques for generating features to train a predictive model to predict a customer lifetime value or churn rate. In one embodiment, a method is disclosed comprising receiving a record that includes a plurality of fields and selecting a value associated with a selected field in the plurality of fields. The method then queries a lookup table comprising a mapping of values to aggregated statistics using the value and receives an aggregated statistic based on the querying. Next, the method generates a feature vector by annotating the record with the aggregated statistic. The method uses this feature vector as an input to a predictive model.

Abstract: Examples of simulated powdered model generation for neural networks are described herein. In some examples, a set of simulated powdered models that simulate application of powder to a computer aided design (CAD) model are generated. In some examples, synthetic images of the simulated powdered models placed in a virtual environment are generated. In some examples, a neural network is trained to recognize real objects based on the synthetic images of the simulated powdered models.

Abstract: In one example in accordance with the present disclosure, a system is described. The system includes a reader to extract cleaning instructions associated with a three-dimensional (3D) printed object. The cleaning instructions include a termination condition to indicate when object cleaning is complete. The system also includes a controller to instruct at least one cleaning device to clean the 3D printed object based on the cleaning instructions. A measurement system of the system determines when the termination condition is met.

Abstract: The invention relates to a process for preparing a dicarboxylic acid comprising the steps: (a) heating a lactone in the presence of a first catalyst system to produce an alkenoic acid; and (b) contacting the alkenoic acid with carbon monoxide, water and a second catalyst system to produce a reaction composition comprising the second catalyst system and the dicarboxylic acid.

Abstract: Disclosed are various embodiments of methods and corresponding devices for effecting such methods that permit integrated circuits, sensors, chips or dies to be identified. Imperfections or irregularities in pixels or memory cells are used to generate identification codes for integrated circuits, sensors, chips or dies. Addresses or data locations of selected defective pixels or memory cells may be used to generate such identification codes.

Abstract: A method for compensating for dark current in an image sensor array. In a representative embodiments, the method includes determining a nominal average dark current for the image sensor array, determining location of each pixel in the image sensor array, obtaining a nominal dark current associated with each pixel based on the nominal average dark current and on the location of the pixel, and subtracting the associated nominal dark current from the image signal for each pixel. At least two of the pixels have differing associated nominal dark currents. In other representative embodiments compensation values for dark currents and for differences in channel processing are determined during the same time period.

Abstract: A system and method for generating dynamically modifiable timing control signals. Timing waveforms can be changed, for each phase of control, by changing the contents of memory locations and start/stop indices. Parallel signals are generated from the output of the memory locations. The memory locations are accessed in a user-selected order by an index value. This technique of generating timing waveforms, and thus control signals, can be applied to, for example, pixel array development.