Abstract: In some aspects, a method includes obtaining feature importance data associated with an alert and indicating relative importance of each of multiple sensor devices and of one or more simulated features. The method includes identifying a group of the sensor devices that have greater relative importance than a highest relative importance of any of the one or more simulated features. In some aspects, a method includes obtaining a reference list of alerts that are similar to a reference alert and a list of alerts predicted to be similar to the reference alert and ranked by predicted similarity to the reference alert. The method includes determining a score indicating similarity of the list to the reference list. A contribution of each alert in the list to the score is determined based on whether that alert appears in the reference list and the rank of that alert in the list.

Abstract: A method includes obtaining a query in a base language and translating the query to generate one or more translated queries each in a respective target language. The method also includes searching one or more sets of electronic files based on the one or more translated queries to generate target-language search results, where each translated query is used to search one or more electronic files that include content in the respective target language of the translated query. The method also includes, based on the target-language search results, scheduling one or more electronic files of the one or more sets of electronic files for at least partial translation to the base language.

Abstract: Calculating energy loss during an outage, including: determining that windspeed data indicating device windspeeds measured at an energy generating device are unavailable within a particular time duration; receiving meteorological data associated with a site location of the energy generating device, the meteorological data including meteorological windspeed data collected within the particular time duration; and predicting one or more estimated device windspeeds at the energy generating device during the particular time duration based on the meteorological data using a trained model for the energy generating device, the trained model being trained using a machine learning algorithm that utilizes historical meteorological windspeed data associated with the site location collected during a previous time duration and corresponding historical device windspeed data measured at the energy generating device during the previous time duration.

Abstract: A method includes receiving, by a processor, an input data set. The input data set includes a plurality of features. The method includes determining, by the processor, one or more characteristics of the input data set. The method includes, based on the one or more characteristics, adjusting, by the processor, one or more architectural parameters of an automated model generation process. The automated model generation process is configured to generate a plurality of models using a weighted randomization process. The one or more architectural parameters weight the weighted randomization process to adjust a probability of generation of models having particular architectural features. The method further includes executing, by the processor, the automated model generation process to output a mode, the model including data representative of a neural network.

Abstract: A method includes obtaining, at a device, a hierarchical structure representing a graphical layout of content items of an electronic document, the content items including at least text. The method also includes generating a word embedding representing a word of the electronic document. The method further includes determining position information of a location of the word in the electronic document. The method also includes determining a descriptor that indicates a relationship of the location to the hierarchical structure. The method further includes providing input data to a machine learning model to generate a semantic region category label of a semantic region of the electronic document. The semantic region includes the word. The input data includes the word embedding, the position information, and the descriptor.

Abstract: A method includes determining, based at least in part on parameters of a software-defined radio (SDR), waveform data descriptive of an electromagnetic waveform. The method also includes generating feature data based on the waveform data and based on one or more symbols decoded from the electromagnetic waveform. The method further includes providing the feature data as input to a first machine-learning model and initiating a response action based on an output of the first machine-learning model.

Abstract: Calibration of online combustion engines using simulations, including: simulating, on a processor coupled to an engine and based on operation data generated during operation of the engine, operation of the engine; training, based on simulating the operation of the engine, one or more trained models; and generating, based at least on the one or more trained models, calibration data corresponding to one or more electronically controllable components of the engine.

Abstract: Uses of artificial intelligence in battery technology including a method that includes receiving a trained model, receiving sensor data from at least one sensor associated with a battery, and executing the trained model by a processor. Executing the trained model includes providing the sensor data as input to the trained model to generate a model output. The method also includes sending, from the processor to a charge controller coupled to the battery, a control signal that is based on the model output and automatically, by the charge controller, initiating or terminating charging of the battery based on the control signal.

Abstract: A process flow for model-based applications, including: receiving data from one or more data sources; applying at least one first transformation on at least a portion of the data to generate transformed input data encoded according to a predefined format; providing the transformed input data to an executed instance of a model facilitating a prediction associated with the data; and exposing access to application data based on an output associated with the model.

Abstract: Controlling the operation of one or more components in an electric vehicle, including: receiving, from one or more vehicle operation sensors, operation data including sensor data corresponding to a condition of one or more components of the vehicle; determining, using a trained model, whether the one or more components of the vehicle are operating in an acceptable manner; and generating a control signal to adjust operation of the one or more components of the vehicle.

Abstract: A method includes determining, based at least in part on parameters of a software defined radio (SDR), waveform data descriptive of an electromagnetic waveform. The method also includes generating feature data based on the waveform data. The method further includes providing the feature data as input to a first machine learning model to predict a future action of a device associated with at least a portion of the electromagnetic waveform and initiating a response action based on the predicted future action.

Abstract: A method includes receiving, at a first server from a second server, a first file attribute associated with a file. The method includes making a determination, at the first server based on the first file attribute, of availability of a classification for the file from a cache of the first server. The method includes, in response to the determination indicating that the classification is not available from the cache, sending a notification to the second server indicating that the classification for the file is not available. The method also includes receiving a first classification for the file from the second server at the first server. The first classification is generated by the second server responsive to the notification.

Abstract: Translating text encodings of machine learning models to executable code, the method comprising: receiving a text encoding of a machine learning model; generating, based on the text encoding of the machine learning model, compilable code encoding the machine learning model; and generating, based on the compilable code, executable code encoding the machine learning model.

Abstract: A method of detecting deviation from an operational state of a rotational device includes receiving, from one or more sensor devices coupled to the rotational device, frequency domain data indicative of vibration data sensed during a sensing period. The method also includes processing the frequency domain data using a trained anomaly detection model to generate an anomaly score for the sensing period and processing the anomaly score using an alert generation model to determine whether to generate an alert.

Abstract: A method includes generating, by a processor of a computing device, a first plurality of models (including a first number of models) based on a genetic algorithm and corresponding to a first epoch of the genetic algorithm. The method includes determining whether to modify an epoch size for the genetic algorithm during a second epoch of the genetic algorithm based on a convergence metric associated with at least one epoch that is prior to the second epoch. The second epoch is subsequent to the first epoch. The method further includes, based on determining to modify the epoch size, generating a second plurality of models (including a second number of models that is different than the first number) based on the genetic algorithm and corresponding to the second epoch. Each model of the first plurality of models and the second plurality of models includes data representative of neural networks.

Abstract: A method includes selecting a subset of models from a plurality of models. The plurality of models is generated based on a genetic algorithm and corresponds to a first epoch of the genetic algorithm. Each of the plurality of models includes data representative of a neural network. The method includes performing at least one genetic operation of the genetic algorithm with respect to at least one model of the subset to generate a trainable model. The method includes determining a rate of improvement associated with prior backpropagation iterations. The method includes selecting, based on the rate of improvement, one of the trainable model or a prior trainable model as a selected trainable model. The method includes generating the trained model including training the selected trainable model. The method includes adding the trained model as input to a second epoch of the genetic algorithm that is subsequent to the first epoch.

Abstract: A method includes obtaining driver characterization data based on sensor data from one or more sensors onboard a vehicle. The sensor data is captured during a time period that includes multiple discharging operations and multiple recharging operations of a battery of the vehicle. The method also includes providing the driver characterization data as input to a trained model to generate a model output. The model output includes a classification that associates the driver characterization data with a driver type profile. The method also includes generating an alert responsive to a charge of the battery deviating from an expected charge of the battery. The expected charge is based on the classification.

Abstract: A method of detecting deviation from an operational state of a rotational device includes receiving, from one or more sensor devices coupled to the rotational device, frequency domain data indicative of vibration data sensed during a sensing period. The method also includes processing the frequency domain data using a trained anomaly detection model to generate an anomaly score for the sensing period and processing the anomaly score using an alert generation model to determine whether to generate an alert.

Abstract: A method of generating a searchable representation of an electronic document includes obtaining an electronic document specifying a graphical layout of content items including text. The method also includes determining pixel data representing the graphical layout of the content items and providing input data based, at least in part, on the pixel data to a document parsing model. The document parsing model is trained to detect functional regions within the graphical layout based on the input data, assign boundaries to the functional regions based on the input data, and assign a category label to each functional region that is detected. The method also includes matching portions of the text to corresponding functional regions based on the boundaries assigned to the functional regions and locations associated with the portions of the text and storing data representing the content items, the functional regions, and the category labels in a searchable data structure.