Abstract: A system for predicting and summarizing medical events from electronic health records includes a computer memory storing aggregated electronic health records from a multitude of patients of diverse age, health conditions, and demographics including medications, laboratory values, diagnoses, vital signs, and medical notes. The aggregated electronic health records are converted into a single standardized data structure format and ordered arrangement per patient, e.g., into a chronological order. A computer (or computer system) executes one or more deep learning models trained on the aggregated health records to predict one or more future clinical events and summarize pertinent past medical events related to the predicted events on an input electronic health record of a patient having the standardized data structure format and ordered into a chronological order.

Abstract: A method, a system, and a non-transitory computer readable medium for accurate Raman spectroscopy. The method may include executing at least one iteration of the steps of: (i) performing, by an optical measurement system, a calibration process that comprises (a) finding a misalignment between a region of interest defined by a spatial filter, and an impinging beam of radiation that is emitted from an illuminated area of a sample, the impinging beam impinges on the spatial filter; and (b) determining a compensating path of propagation of the impinging beam that compensates the misalignment; and (ii) performing a measurement process, while the optical measurement system is configured to provide the compensating path of propagation of the impinging beam, to provide one or more Raman spectra.

Abstract: In some aspects, the disclosure is directed to methods and systems for monitoring and isolating issues over physical layer (layer 1) paths using Flexible Ethernet (FlexE) protocols via client-based operations, administration, and maintenance (OAM) management. Each node within a FlexE communications path may dynamically insert, extract, and/or replace OAM messages from FlexE data streams as needed to ensure that the OAM messages may be provided unimpeded to a destination (e.g. monitoring server or other device), as well as inserting additional monitoring data (e.g. bit error rates, delay measurements, loss measurements, threshold crossing alerts, or other such monitoring notifications or data) for intermediary nodes.

Abstract: A system for predicting and summarizing medical events from electronic health records includes a computer memory storing aggregated electronic health records from a multitude of patients of diverse age, health conditions, and demographics including medications, laboratory values, diagnoses, vital signs, and medical notes. The aggregated electronic health records are converted into a single standardized data structure format and ordered arrangement per patient, e.g., into a chronological order. A computer (or computer system) executes one or more deep learning models trained on the aggregated health records to predict one or more future clinical events and summarize pertinent past medical events related to the predicted events on an input electronic health record of a patient having the standardized data structure format and ordered into a chronological order.

Abstract: A system for predicting and summarizing medical events from electronic health records includes a computer memory storing aggregated electronic health records from a multitude of patients of diverse age, health conditions, and demographics including medications, laboratory values, diagnoses, vital signs, and medical notes. The aggregated electronic health records are converted into a single standardized data structure format and ordered arrangement per patient, e.g., into a chronological order. A computer (or computer system) executes one or more deep learning models trained on the aggregated health records to predict one or more future clinical events and summarize pertinent past medical events related to the predicted events on an input electronic health record of a patient having the standardized data structure format and ordered into a chronological order.

Abstract: A method is described for training a predictive model which increases the interpretability and trustworthiness of the model for end-users. The model is trained from data having multitude of features. Each feature is associated with a real value and a time component. Many predicates (atomic elements for training the model) are defined as binary functions operating on the features, and typically time sequences of the features or logical combinations thereof. The predicates can be limited to those functions which have human understandability or encode expert knowledge relative to a predication task of the model. We iteratively train a boosting model with input from an operator or human-in-the-loop. The human-in-the-loop is provided with tools to inspect the model as it is iteratively built and remove one or more of the predicates in the model, e.g. if it does not have indicia of trustworthiness, is not causally related to a prediction of the model, or is not understandable.

Abstract: The present disclosure provides systems and methods for estimating audience member engagement with content, or distinguishing between users consuming content and users who have become distracted or have left. User presence or attention may be estimated from user interactions with or skipping of content, with the user interactions either compared to high quality engagement data from small audience measurement panels, or extrapolated based on a temporal-engagement curve. An attention gap may be estimated, representing users that were not present for or not engaged with or paying attention to a presentation of content at a client device. This allows the measurement system to distinguish between users who consumed and potentially enjoyed the content, and users who did not, even as client devices of both sets of users receive and present the content items.

Abstract: A machine learning method is described for generating labels for members of a training set where the labels are not directly available in the training set data. In a first stage of the method an iterative process is used to gradually build up a list of features (“partition features” herein) which are conceptually related to the class label using a human-in-the loop (expert). In a second part of the process we generate labels for the members of the training set, build up a boosting model using the labeling to come up with additional partition features, score the labeling of the training set members from the boosting model, and then with the human-in-the-loop evaluate a labels assigned to a small subset of the members depending on their score. The labels assigned to some or all of those members in the subset may be flipped depending on the evaluation. The final outcome of the process is an interpretable model that explains how the labels were generated and a labeled set of training data.

Abstract: In some aspects, the disclosure is directed to methods and systems for monitoring and isolating issues over physical layer (layer 1) paths using Flexible Ethernet (FlexE) protocols via client-based operations, administration, and maintenance (OAM) management. Each node within a FlexE communications path may dynamically insert, extract, and/or replace OAM messages from FlexE data streams as needed to ensure that the OAM messages may be provided unimpeded to a destination (e.g. monitoring server or other device), as well as inserting additional monitoring data (e.g. bit error rates, delay measurements, loss measurements, threshold crossing alerts, or other such monitoring notifications or data) for intermediary nodes.

Abstract: The present disclosure provides systems and methods for estimating audience member engagement with content, or distinguishing between users consuming content and users who have become distracted or have left. User presence or attention may be estimated from user interactions with or skipping of content, with the user interactions either compared to high quality engagement data from small audience measurement panels, or extrapolated based on a temporal-engagement curve. An attention gap may be estimated, representing users that were not present for or not engaged with or paying attention to a presentation of content at a client device. This allows the measurement system to distinguish between users who consumed and potentially enjoyed the content, and users who did not, even as client devices of both sets of users receive and present the content items.

Abstract: The present disclosure provides systems and methods for estimating audience member engagement with content, or distinguishing between users consuming content and users who have become distracted or have left. User presence or attention may be estimated from user interactions with or skipping of content, with the user interactions either compared to high quality engagement data from small audience measurement panels, or extrapolated based on a temporal-engagement curve. An attention gap may be estimated, representing users that were not present for or not engaged with or paying attention to a presentation of content at a client device. This allows the measurement system to distinguish between users who consumed and potentially enjoyed the content, and users who did not, even as client devices of both sets of users receive and present the content items.

Abstract: The present disclosure provides systems and methods for estimating audience member engagement with content, or distinguishing between users consuming content and users who have become distracted or have left. User presence or attention may be estimated from user interactions with or skipping of content, with the user interactions either compared to high quality engagement data from small audience measurement panels, or extrapolated based on a temporal-engagement curve. An attention gap may be estimated, representing users that were not present for or not engaged with or paying attention to a presentation of content at a client device. This allows the measurement system to distinguish between users who consumed and potentially enjoyed the content, and users who did not, even as client devices of both sets of users receive and present the content items.

Abstract: A system for predicting and summarizing medical events from electronic health records includes a computer memory storing aggregated electronic health records from a multitude of patients of diverse age, health conditions, and demographics including medications, laboratory values, diagnoses, vital signs, and medical notes. The aggregated electronic health records are converted into a single standardized data structure format and ordered arrangement per patient, e.g., into a chronological order. A computer (or computer system) executes one or more deep learning models trained on the aggregated health records to predict one or more future clinical events and summarize pertinent past medical events related to the predicted events on an input electronic health record of a patient having the standardized data structure format and ordered into a chronological order.

Abstract: A system for predicting and summarizing medical events from electronic health records includes a computer memory storing aggregated electronic health records from a multitude of patients of diverse age, health conditions, and demographics including medications, laboratory values, diagnoses, vital signs, and medical notes. The aggregated electronic health records are converted into a single standardized data structure format and ordered arrangement per patient, e.g., into a chronological order. A computer (or computer system) executes one or more deep learning models trained on the aggregated health records to predict one or more future clinical events and summarize pertinent past medical events related to the predicted events on an input electronic health record of a patient having the standardized data structure format and ordered into a chronological order.

Abstract: A system for predicting and summarizing medical events from electronic health records includes a computer memory storing aggregated electronic health records from a multitude of patients of diverse age, health conditions, and demographics including medications, laboratory values, diagnoses, vital signs, and medical notes. The aggregated electronic health records are converted into a single standardized data structure format and ordered arrangement per patient, e.g., into a chronological order. A computer (or computer system) executes one or more deep learning models trained on the aggregated health records to predict one or more future clinical events and summarize pertinent past medical events related to the predicted events on an input electronic health record of a patient having the standardized data structure format and ordered into a chronological order.

Abstract: The present disclosure provides systems and methods for estimating audience member engagement with content, or distinguishing between users consuming content and users who have become distracted or have left. User presence or attention may be estimated from user interactions with or skipping of content, with the user interactions either compared to high quality engagement data from small audience measurement panels, or extrapolated based on a temporal-engagement curve. An attention gap may be estimated, representing users that were not present for or not engaged with or paying attention to a presentation of content at a client device. This allows the measurement system to distinguish between users who consumed and potentially enjoyed the content, and users who did not, even as client devices of both sets of users receive and present the content items.

Abstract: A packet switching system for a packet transfer network, the system having an architecture including a plurality of line cards, each including an ingress path pipeline, with processing elements, and an egress buffer, and an electro-optical In/Out (IO) interconnect coupling the line cards to one another in a full mesh connectivity, in the absence of a switch fabric, wherein the ingress path pipeline of each line card is coupled by means of the electro-optical IO interconnect to the egress buffer of each of the plurality of line cards.

Abstract: A method of error mitigation for transferring packets over a chip-to-chip data interconnect using a high speed interconnect protocol, the method including grouping a pre-selected number of high speed interconnect protocol words to form a protection frame, adding at least one additional error protection bit to each word in the group, adding a synchronization bit to each word, using the synchronization bit in a first word in each frame for synchronization of the protection frame and detecting and correcting a single bit error in the protection frame using the additional error protection bits, thereby reducing packet drop when the frames are transferred over the high speed data interconnect.

Abstract: An architecture for a line card in a network routing device is provided. The line card architecture provides a bi-directional interface between the routing device and a network, both receiving packets from the network and transmitting the packets to the network through one or more connecting ports. In both the receive and transmit path, packets processing and routing in a multi-stage, parallel pipeline that can operate on several packets at the same time to determine each packet's routing destination is provided. Once a routing destination determination is made, the line card architecture provides for each received packet to be modified to contain new routing information and additional header data to facilitate packet transmission through the switching fabric. The line card architecture further provides for the use of bandwidth management techniques in order to buffer and enqueue each packet for transmission through the switching fabric to a corresponding destination port.

Abstract: A packet switching system for a packet transfer network, the system having an architecture including a plurality of line cards, each including an ingress path pipeline, with processing elements, and an egress buffer, and an electro-optical In/Out (IO) interconnect coupling the line cards to one another in a full mesh connectivity, in the absence of a switch fabric, wherein the ingress path pipeline of each line card is coupled by means of the electro-optical IO interconnect to the egress buffer of each of the plurality of line cards.