Abstract: A computer implemented method for incorporating multiple objectives in a ranked list of search results includes receiving a search query from a client device, accessing a set of stored listings for goods or services and probabilities of serving the listings, defining a serving vector as a probability distribution over the set of listings, providing a serving vector as input to a multi-objective function, decomposing the multi-objective function into one or more objective functions, generating a ranked list of the listings based at least in part on the serving vector that maximizes the decomposed multi-objective function, and providing the listings to the client device according to the order of the ranked list. Each objective function addresses a different goal in an overall diversity optimization.

Abstract: A system uses machine models to estimate trip durations or distance. The system trains a historical model to estimate trip duration using characteristics of past trips. The system trains a real-time model to estimate trip duration using characteristics of recently completed trips. The historical and real-time models may use different time windows of training data to predict estimates, and may be trained to predict an adjustment to an initial trip estimate. A selector model is trained to predict whether the historical model, the real-time model, or a combination of the historical and real-time models will more accurately estimate a trip duration, given features associated with a trip duration request, and the system accordingly uses the models to estimate a trip duration. In some embodiments, the real-time model and the selector may be trained using batch machine learning techniques which allow the models to incorporate new trip data as trips complete.

Abstract: A system uses machine models to estimate trip durations or distance. The system trains a historical model to estimate trip duration using characteristics of past trips. The system trains a real-time model to estimate trip duration using characteristics of recently completed trips. The historical and real-time models may use different time windows of training data to predict estimates, and may be trained to predict an adjustment to an initial trip estimate. A selector model is trained to predict whether the historical model, the real-time model, or a combination of the historical and real-time models will more accurately estimate a trip duration, given features associated with a trip duration request, and the system accordingly uses the models to estimate a trip duration. In some embodiments, the real-time model and the selector may be trained using batch machine learning techniques which allow the models to incorporate new trip data as trips complete.

Abstract: A computer implemented method for incorporating multiple objectives in a ranked list of search results includes receiving a search query from a client device, accessing a set of stored listings for goods or services and probabilities of serving the listings, defining a serving vector as a probability distribution over the set of listings, providing a serving vector as input to a multi-objective function, decomposing the multi-objective function into one or more objective functions, generating a ranked list of the listings based at least in part on the serving vector that maximizes the decomposed multi-objective function, and providing the listings to the client device according to the order of the ranked list. Each objective function addresses a different goal in an overall diversity optimization.

Abstract: A system uses machine models to estimate trip durations or distance. The system trains a historical model to estimate trip duration using characteristics of past trips. The system trains a real-time model to estimate trip duration using characteristics of recently completed trips. The historical and real-time models may use different time windows of training data to predict estimates, and may be trained to predict an adjustment to an initial trip estimate. A selector model is trained to predict whether the historical model, the real-time model, or a combination of the historical and real-time models will more accurately estimate a trip duration, given features associated with a trip duration request, and the system accordingly uses the models to estimate a trip duration. In some embodiments, the real-time model and the selector may be trained using batch machine learning techniques which allow the models to incorporate new trip data as trips complete.

Abstract: A transport service system determines the accuracy of a map matched trajectory using a forward probability algorithm. A transport vehicle on a trip relays location data to the system. The system uses a map of the corresponding area and the location data to calculate an emission probability, the likelihood of a candidate road being associated with a location data point, and a transition probability, the likelihood of a second state occurring after a first state. The joint probability of the emission and transition probabilities is used to determine a total number of zero forward probability occurrences and an average forward probability associated with the trip. These metrics are used to measure the accuracy of the map matching algorithm for the trip.

Abstract: A transport service system determines the accuracy of a map matched trajectory using a forward probability algorithm. A transport vehicle on a trip relays location data to the system. The system uses a map of the corresponding area and the location data to calculate an emission probability, the likelihood of a candidate road being associated with a location data point, and a transition probability, the likelihood of a second state occurring after a first state. The joint probability of the emission and transition probabilities is used to determine a total number of zero forward probability occurrences and an average forward probability associated with the trip. These metrics are used to measure the accuracy of the map matching algorithm for the trip.

Abstract: A system uses machine models to estimate trip durations or distance. The system trains a historical model to estimate trip duration using characteristics of past trips. The system trains a real-time model to estimate trip duration using characteristics of recently completed trips. The historical and real-time models may use different time windows of training data to predict estimates, and may be trained to predict an adjustment to an initial trip estimate. A selector model is trained to predict whether the historical model, the real-time model, or a combination of the historical and real-time models will more accurately estimate a trip duration, given features associated with a trip duration request, and the system accordingly uses the models to estimate a trip duration. In some embodiments, the real-time model and the selector may be trained using batch machine learning techniques which allow the models to incorporate new trip data as trips complete.

Abstract: A system uses machine models to estimate trip durations or distance. The system trains a historical model to estimate trip duration using characteristics of past trips. The system trains a real-time model to estimate trip duration using characteristics of recently completed trips. The historical and real-time models may use different time windows of training data to predict estimates, and may be trained to predict an adjustment to an initial trip estimate. A selector model is trained to predict whether the historical model, the real-time model, or a combination of the historical and real-time models will more accurately estimate a trip duration, given features associated with a trip duration request, and the system accordingly uses the models to estimate a trip duration. In some embodiments, the real-time model and the selector may be trained using batch machine learning techniques which allow the models to incorporate new trip data as trips complete.

Abstract: A method of reading image data from an image sensor includes accumulating image charges in photosensitive elements of an array of pixel cells. The accumulated image charges are transferred to corresponding transistors in multi-phase transfer channels that are coupled to corresponding columns of the pixel array. Multi-phase transfer signals are generated. Each set of the multi-phase transfer signals includes a plurality of control signals that are out-of-phase with one another and are coupled to control respective transistors in the multi-phase transfer channels. The accumulated image charges from a first variable number of pixel cells of a selected column are output in response to the multi-phase transfer signals. The accumulated image charges from a second variable number of pixel cells of another selected column are output in response to the multi-phase transfer signals.

Abstract: A method of reading image data from an image sensor includes accumulating image charges in photosensitive elements of an array of pixel cells. The accumulated image charges are transferred to corresponding transistors in multi-phase transfer channels that are coupled to corresponding columns of the pixel array. Multi-phase transfer signals are generated. Each set of the multi-phase transfer signals includes a plurality of control signals that are out-of-phase with one another and are coupled to control respective transistors in the multi-phase transfer channels. The accumulated image charges from a first variable number of pixel cells of a selected column are output in response to the multi-phase transfer signals. The accumulated image charges from a second variable number of pixel cells of another selected column are output in response to the multi-phase transfer signals.