Abstract: The present disclosure relates generally to improving calorie expenditure prediction and tracking and, more particularly, to techniques for calibration and calorimetry using data from motions sensors and heart rate sensors. Embodiments of the present disclosure include a fitness tracking device and techniques for accurately tracking an individual's energy expenditure over time and over a variety of activities while wearing the fitness tracking device. In some embodiments, the fitness tracking device may be a wearable device. The wearable device may be worn on a wrist, such as a watch, and it may include one or more microprocessors, a display, and a variety of sensors, including a heart rate sensor and one or more motion sensors.

Abstract: In one aspect, the present disclosure relates to a method including obtaining a plurality of heart rate measurements of the user over a period of time; obtaining motion data of the user over the period of time; analyzing the motion data of the user to determine for each of the plurality of heart rate measurements, a corresponding work rate measurement; determining, for each of the plurality of heart rate measurements, a first confidence level; determining, for each corresponding work rate measurement, a second confidence level; and estimating a first energy expenditure rate using the plurality of heart rate measurements; estimating a second energy expenditure rate using the plurality of work rate measurements; and estimating a weighted energy expenditure rate of the user by combining the first energy expenditure rate weighted by the first confidence level and the second energy expenditure rate weighted by the second confidence level.

Abstract: In one aspect, the present disclosure relates to a method including obtaining, by a heart rate sensor of a fitness tracking device, a heart rate measurement of a user of the fitness tracking device; obtaining, by at least one motion sensor, motion data of the user; analyzing, by the fitness tracking device, the motion data of the user to estimate a step rate of the user; estimating, by the fitness tracking device, a load associated with a physical activity of the user by comparing the heart rate measurement with the step rate of the user; and estimating, by the fitness tracking device, an energy expenditure rate of the user using the load and at least one of the heart rate measurement and the step rate.

Abstract: In one aspect, the present disclosure relates to a method, including obtaining, by the fitness tracking device, motion data of the user over a period of time, wherein the motion data can include a first plurality motion measurements from a first motion sensor of the fitness tracking device; determining, by the fitness tracking device, using the motion data an angle of the fitness tracking device relative to a plane during the period of time; estimating by the fitness tracking device, using the motion data, a range of linear motion of the fitness tracking device through space during the period of time; and comparing, by the fitness tracking device, the angle of the fitness tracking device to a threshold angle and comparing the range of linear motion of the fitness tracking device to a threshold range of linear motion to determine whether the user is sitting or standing.

Abstract: An example method for anomaly detection in streaming data includes applying statistical analysis to streaming data in a sliding window. The method also includes extracting a feature. The method also includes determining class assignment for the feature using class conditional probability densities and a threshold.

Abstract: A method of determining response similarity neighborhoods comprises extracting data and spatial locations from a number of nodes, and with a processor, time aligning data traces, computing a feature vector of the extracted data, defining a neighborhood of the nodes, and determining similarities between a target node and a number of neighbor nodes within the neighborhood of the target node.

Abstract: Methods, systems, and computer-readable and executable instructions are provided for classifying an electrocardiogram (ECG) signal. Classifying an ECG signal can include analyzing the ECG signal using a stream of pulses generated by a sampler, extracting cardiac pulse features from a timing of the stream of pulses, and classifying the ECG signal based on the extracted cardiac pulse feature.

Abstract: Systems and methods for multi-regime detection in streaming data are disclosed. An example method includes generating vectors for each sample of the streaming data. The method also includes inducing mean independence of the vectors to find an embedded data trajectory. The method also includes comparing the embedded data trajectory with known data trajectories. The method also includes issuing an alert if the embedded data trajectory corresponds to a known data trajectory indicating an anomaly in the streaming data.

Abstract: Methods, systems, and computer-readable and executable instructions are provided for classifying an electrocardiogram (ECG) signal. Classifying an ECG signal can include analyzing the ECG signal using a stream of pulses generated by a sampler, extracting cardiac pulse features from a timing of the stream of pulses, and classifying the ECG signal based on the extracted cardiac pulse feature.

Abstract: Systems and methods for anomaly detection in streaming data are disclosed. An example method includes applying statistical analysis to streaming data in a sliding window. The method also includes extracting a feature. The method also includes determining class assignment for the feature using class conditional probability densities and a threshold.

Abstract: Systems and methods for multi-regime detection in streaming data are disclosed. An example method includes generating vectors for each sample of the streaming data. The method also includes inducing mean independence of the vectors to find an embedded data trajectory. The method also includes comparing the embedded data trajectory with known data trajectories. The method also includes issuing an alert if the embedded data trajectory corresponds to a known data trajectory indicating an anomaly in the streaming data.

Abstract: Systems and methods of time encoding using an integrate and fire (IF) sampler are disclosed. In an example, a method includes receiving input signals for separate classes. The method also includes generating a pulse train based on the input signals. The method also includes binning the pulse train to generate a feature vector.