Abstract: The disclosure herein describes platform-level checkpointing for deep learning (DL) jobs. The checkpointing is performed through capturing two kinds of state data: (i) GPU state (device state), and (ii) CPU state (host state). The GPU state includes GPU data (e.g., model parameters, optimizer state, etc.) that is located in the GPU and GPU context (e.g., the default stream in GPU, various handles created by the libraries such as DNN, Blas, etc.). Only a fraction of the GPU memory is copied because the checkpointing is done in a domain-aware manner. The “active” memory contains useful data like model parameters. To be able to capture the useful data, memory management is controlled to identify which parts of the memory are active. Also, to restore the destination GPU to the same context/state, a mechanism is used to capture such state-changing events on an original GPU and replayed on a destination GPU.

Abstract: A system and method are disclosed for aggregating product trends. Embodiments include receiving an initial set of images from one or more data feeds by a trend aggregation system comprising a server, identifying one or more attribute values from the initial set of images, and quantifying a social affinity score of one or more items of an inventory of a supply chain entity based on recentness, relevance, and similarities of the identified one or more attribute values to an attribute value of a potential product for a product assortment. Embodiments may further include receiving an image of at least one additional item, identifying a product attribute from the image, and assigning an attribute value to the at least one additional item based, at least in part, on the identified attribute value from the image of the least one additional item and the attribute values of the at least two items.

Abstract: Aspects of the present disclosure provide techniques for classifying a trip. Embodiments include receiving, from a plurality of users, a plurality of historical trip records. Each of the plurality of historical trip records may comprise one or more historical trip attributes and historical classification information. Embodiments include training a predictive model, using the plurality of historical trip records, to classify trips based on trip records. Training the predictive model may comprise determining a plurality of hot spots based on the historical trip records, each of the plurality of hot spots comprising a region encompassing one or more locations, and associating, in the predictive model, the plurality of hot spots with historical classification information. Embodiments include receiving, from a user, a new trip record comprising a plurality of trip attributes related to a trip and using the predictive model to predict a classification for the trip based on the trip record.

Abstract: Systems and methods for data visualization in the metaverse with portability to multiple metaverse channels are disclosed. In one embodiment, a method for data visualization in the metaverse with portability to multiple metaverse channels may include: (1) ingesting, by a data rendering computer program, data from a plurality of data sources, each data source associated with an entity; (2) categorizing, by the data rendering computer program, the ingested data into a plurality of categories, wherein each category of data comprises data from a subset of two or more of the plurality of the data sources; and (3) streaming, by the data rendering computer program, one of the categories of data to an immersive input/output device associated with a user over one of a plurality of metaverse channels, wherein the category of data is consumed by the immersive input/output device.

Abstract: Methods and systems for automatically and dynamically ingesting healthcare data from a plurality of data sources into a single unified ingestion database, in a real-time include a plurality of client computing devices for capturing healthcare data of a patient. A central server is connected to each client computing device. A memory of the server includes one or more ingestion data tables to store healthcare data received from the client computers and/or a plurality of data sources in a predetermined data-pattern. An ingestion module includes algorithms to process the input healthcare data. The central server receives the input healthcare data from the client computing devices and the ingestion module identifies an ingestion data-pattern for the ingestion data table and corresponding mapping codes, in accordance with the historical healthcare data received from the plurality of data-sources. The input healthcare data is processed and ingested within the ingestion table using the mapping codes.

Abstract: The disclosure herein describes platform-level checkpointing for deep learning (DL) jobs. The checkpointing is performed through capturing two kinds of state data: (i) GPU state (device state), and (ii) CPU state (host state). The GPU state includes GPU data (e.g., model parameters, optimizer state, etc.) that is located in the GPU and GPU context (e.g., the default stream in GPU, various handles created by the libraries such as DNN, Blas, etc.). Only a fraction of the GPU memory is copied because the checkpointing is done in a domain-aware manner. The “active” memory contains useful data like model parameters. To be able to capture the useful data, memory management is controlled to identify which parts of the memory are active. Also, to restore the destination GPU to the same context/state, a mechanism is used to capture such state-changing events on an original GPU and replayed on a destination GPU.

Abstract: Aspects of the present disclosure provide techniques for classifying a trip. Embodiments include receiving, from a plurality of users, a plurality of historical trip records. Each of the plurality of historical trip records may comprise one or more historical trip attributes and historical classification information. Embodiments include training a predictive model, using the plurality of historical trip records, to classify trips based on trip records. Training the predictive model may comprise determining a plurality of hot spots based on the historical trip records, each of the plurality of hot spots comprising a region encompassing one or more locations, and associating, in the predictive model, the plurality of hot spots with historical classification information. Embodiments include receiving, from a user, a new trip record comprising a plurality of trip attributes related to a trip and using the predictive model to predict a classification for the trip based on the trip record.

Abstract: Aspects of the present disclosure provide techniques for classifying a trip. Embodiments include receiving, from a plurality of users, a plurality of historical trip records. Each of the plurality of historical trip records may comprise one or more historical trip attributes and historical classification information. Embodiments include training a predictive model, using the plurality of historical trip records, to classify trips based on trip records. Training the predictive model may comprise determining a plurality of hot spots based on the historical trip records, each of the plurality of hot spots comprising a region encompassing one or more locations, and associating, in the predictive model, the plurality of hot spots with historical classification information. Embodiments include receiving, from a user, a new trip record comprising a plurality of trip attributes related to a trip and using the predictive model to predict a classification for the trip based on the trip record.

Abstract: Aspects of the present disclosure provide techniques for classifying a trip. Embodiments include receiving, from a plurality of users, a plurality of historical trip records. Each of the plurality of historical trip records may comprise one or more historical trip attributes and historical classification information. Embodiments include training a predictive model, using the plurality of historical trip records, to classify trips based on trip records. Training the predictive model may comprise determining a plurality of hot spots based on the historical trip records, each of the plurality of hot spots comprising a region encompassing one or more locations, and associating, in the predictive model, the plurality of hot spots with historical classification information. Embodiments include receiving, from a user, a new trip record comprising a plurality of trip attributes related to a trip and using the predictive model to predict a classification for the trip based on the trip record.

Abstract: Certain aspects of the present disclosure provide techniques for intelligent grouping of travel data for review through a user interface. In one embodiment, a method for providing grouped travel data to a user interface of an application, comprises: receiving a plurality of trip records from an application running on a remote device; providing a first subset of the plurality of trip records to a prediction model; providing a second subset of the plurality of trip records to a model training module; receiving labels for each trip record of the first subset of the plurality of trip records from the prediction model; grouping the first subset of the plurality of trip records based on the received labels; and transmitting the grouped first subset of the plurality of trip records to the application running on the remote device.

Abstract: The disclosure herein describes processing deletion requests using sequencing numbers with change feed updates. When a deletion occurs on the source data store, a deletion notification is created in a change feed on the source server. The deletion notification includes a set of deletion record IDs identifying a set of records to be deleted, a tombstone sequence number (TSN) identifying a sequence of the deletion notification within a set of deletion notifications and/or a deletion sequence number (DSN). The DSN is incremented by one each time a new deletion notification is created. A deletion notification can represent deletion of a single record or a set of records. Each deletion notification is assigned a time-to-live (TTL) value. The deletion notification is deleted at expiration of the TTL. The TSN and the DSN entries are used to determine whether any deletion updates have been missed to prevent silent failures.

Abstract: A system and method are provided for managing creation of data objects, for example in a network or big data environment. A method comprises: receiving, at a processor, a data object creation command for creating the data object; comparing, at the processor, at least one data object creation parameter with a stored data object creation standard; and initiating creation of the data object in response to the data object creation parameter matching a corresponding parameter in the stored data object creation standard. The data object creation command can be modified to meet the stored data object creation standard, and to change an associated permission level. Embodiments of the present disclosure encapsulate standards defining how to create a data object, such that it is no longer necessary to have a person to make sure that these standards are followed, and automate management of object creation while ensuring conformity to organizational standards.

Abstract: Aspects of the present disclosure provide techniques for classifying a trip. Embodiments include receiving, from a plurality of users, a plurality of historical trip records. Each of the plurality of historical trip records may comprise one or more historical trip attributes and historical classification information. Embodiments include training a predictive model, using the plurality of historical trip records, to classify trips based on trip records. Training the predictive model may comprise determining a plurality of hot spots based on the historical trip records, each of the plurality of hot spots comprising a region encompassing one or more locations, and associating, in the predictive model, the plurality of hot spots with historical classification information. Embodiments include receiving, from a user, a new trip record comprising a plurality of trip attributes related to a trip and using the predictive model to predict a classification for the trip based on the trip record.

Abstract: Method and system to extract domain concepts to create domain dictionaries and ontologies comprises collecting a plurality of reference papers and further classifying the collected plurality of reference papers as relevant and irrelevant. Each of the ‘relevant’ reference papers is further processed by the system, during which the system identifies relevant sections from each document and further processes data in the relevant sections to extract required information and also to identify a relationship between different extracted information, which is further used to create domain dictionaries and ontologies.

Abstract: The disclosure herein describes processing deletion requests using sequencing numbers with change feed updates. When a deletion occurs on the source data store, a deletion notification is created in a change feed on the source server. The deletion notification includes a set of deletion record IDs identifying a set of records to be deleted, a tombstone sequence number (TSN) identifying a sequence of the deletion notification within a set of deletion notifications and/or a deletion sequence number (DSN). The DSN is incremented by one each time a new deletion notification is created. A deletion notification can represent deletion of a single record or a set of records. Each deletion notification is assigned a time-to-live (TTL) value. The deletion notification is deleted at expiration of the TTL. The TSN and the DSN entries are used to determine whether any deletion updates have been missed to prevent silent failures.

Abstract: The disclosure herein describes platform-level checkpointing for deep learning (DL) jobs. The checkpointing is performed through capturing two kinds of state data: (i) GPU state (device state), and (ii) CPU state (host state). The GPU state includes GPU data (e.g., model parameters, optimizer state, etc.) that is located in the GPU and GPU context (e.g., the default stream in GPU, various handles created by the libraries such as DNN, Blas, etc.). Only a fraction of the GPU memory is copied because the checkpointing is done in a domain-aware manner. The “active” memory contains useful data like model parameters. To be able to capture the useful data, memory management is controlled to identify which parts of the memory are active. Also, to restore the destination GPU to the same context/state, a mechanism is used to capture such state-changing events on an original GPU and replayed on a destination GPU.

Abstract: The disclosure herein describes processing deletion requests using sequencing numbers with change feed updates. When a deletion occurs on the source data store, a deletion notification is created in a change feed on the source server. The deletion notification includes a set of deletion record IDs identifying a set of records to be deleted, a tombstone sequence number (TSN) identifying a sequence of the deletion notification within a set of deletion notifications and/or a deletion sequence number (DSN). The DSN is incremented by one each time a new deletion notification is created. A deletion notification can represent deletion of a single record or a set of records. Each deletion notification is assigned a time-to-live (TTL) value. The deletion notification is deleted at expiration of the TTL. The TSN and the DSN entries are used to determine whether any deletion updates have been missed to prevent silent failures.

Abstract: Aspects of the present disclosure provide techniques for classifying a trip. Embodiments include receiving, from a plurality of users, a plurality of historical trip records. Each of the plurality of historical trip records may comprise one or more historical trip attributes and historical classification information. Embodiments include training a predictive model, using the plurality of historical trip records, to classify trips based on trip records. Training the predictive model may comprise determining a plurality of hot spots based on the historical trip records, each of the plurality of hot spots comprising a region encompassing one or more locations, and associating, in the predictive model, the plurality of hot spots with historical classification information. Embodiments include receiving, from a user, a new trip record comprising a plurality of trip attributes related to a trip and using the predictive model to predict a classification for the trip based on the trip record.

Abstract: Aspects of the present disclosure provide techniques for classifying a trip. Embodiments include receiving, from a plurality of users, a plurality of historical trip records. Each of the plurality of historical trip records may comprise one or more historical trip attributes and historical classification information. Embodiments include training a predictive model, using the plurality of historical trip records, to classify trips based on trip records. Training the predictive model may comprise determining a plurality of hot spots based on the historical trip records, each of the plurality of hot spots comprising a region encompassing one or more locations, and associating, in the predictive model, the plurality of hot spots with historical classification information. Embodiments include receiving, from a user, a new trip record comprising a plurality of trip attributes related to a trip and using the predictive model to predict a classification for the trip based on the trip record.

Abstract: The disclosure herein describes processing deletion requests using sequencing numbers with change feed updates. When a deletion occurs on the source data store, a deletion notification is created in a change feed on the source server. The deletion notification includes a set of deletion record IDs identifying a set of records to be deleted, a tombstone sequence number (TSN) identifying a sequence of the deletion notification within a set of deletion notifications and/or a deletion sequence number (DSN). The DSN is incremented by one each time a new deletion notification is created. A deletion notification can represent deletion of a single record or a set of records. Each deletion notification is assigned a time-to-live (TTL) value. The deletion notification is deleted at expiration of the TTL. The TSN and the DSN entries are used to determine whether any deletion updates have been missed to prevent silent failures.