Google Inc.

Facebook, Inc.

Equatine Labs, Inc.

Joshua A Redstone

Tushar Deepak Chandra

Yoram Singer

Tal Shaked

Eyal Michael Sharon

Tze Way Eugene Ie

Srinivasa Narayanan

Michael Burrows

Sean Quinlan

Jonathan D Chang

Mukund Narasimhan

Yuntao Jia

Daniel Jeng-Ping Hui

Tze Way Eugene Le

Benjamin J Gertzfield

The present disclosure provides systems and techniques for efficient locking of datasets in a database when updates to a dataset may be delayed. A method may include accumulating a plurality of updates to a first set of one or more values associated with one or more features. The first set of one or more values may be stored within a first database column. Next, it may be determined that a first database column update aggregation rule is satisfied. A lock assigned to at least a portion of at least a first database column may be acquired. Accordingly, one or more values in the first set within the first database column may be updated based on the plurality of updates. In an implementation, the first set of one or more values may be associated with the first lock.

Efficient locking of large data collections

In one embodiment, a computing device accesses a social graph comprising nodes and edges connecting the nodes. Each of the edges between two of the nodes represents a single degree of separation between them. The nodes include user nodes corresponding to users of an online social network, and concept nodes corresponding to places. A particular place corresponds to a particular concept node, and corresponds to an associated location and a perimeter. A number of check-ins are received, each check-in being associated with the particular place and having a geographic location. Each check-in corresponds to an edge of the social graph between a user node and the particular concept node. Based on the received check-ins the computing device determines whether to update the associated location and the perimeter.

Determining a location and area of a place

The present disclosure provides methods and systems for using variable length representations of machine learning statistics. A method may include storing an n-bit representation of a first statistic at a first n-bit storage cell. A first update to the first statistic may be received, and it may be determined that the first update causes a first loss of precision of the first statistic as stored in the first n-bit storage cell. Accordingly, an m-bit representation of the first statistic may be stored at a first m-bit storage cell based on the determination. The first m-bit storage cell may be associated with the first n-bit storage cell. As a result, upon receiving an instruction to use the first statistic in a calculation, a combination of the n-bit representation and the m-bit representation may be used to perform the calculation.

Using variable length representations for machine learning statistics

In one embodiment, a user of a social networking system requests to check in a place near the user's current location. The social networking system generates a list of places near the user's current location, ranks the places in the list of places near the user's current location by a distance between each place and the user's current location, as well as activity of the user and the user's social contacts for each place, and returns the ranked list to the user.

Location ranking using social graph information

Methods and systems for replacing feature values of features in training data with integer values selected based on a ranking of the feature values. The methods and systems are suitable for preprocessing large-scale machine learning training data.

Using an integerized representation for large-scale machine learning data

In one embodiment, a method includes calculating a first mean of check-in locations associated with a place; selecting a subset of the check-in locations based on distances between the first mean and the check-in locations; and determining a central location and at least a portion of a perimeter of the place based on the subset of the check-in locations.

Determining a location and area of a place based on distances between the first mean and check in locations

Implementations of the disclosed subject matter provide methods and systems for using a multistage learner for efficiently boosting large datasets in a machine learning system. A method may include obtaining a first plurality of examples for a machine learning system and selecting a first point in time. Next, a second point in time occurring subsequent to the first point in time may be selected. The machine learning system may be trained using m of the first plurality of examples. Each of the m examples may include a feature initially occurring after the second point in time. In addition, the machine learning system may be trained using n of the first plurality of examples, and each of the n examples may include a feature initially occurring after the first point in time.

Multistage learner for efficiently boosting large datasets

The disclosure provides a volumetric measurement device that includes of a vessel into which contents to be measured are placed. The device has a target operational range of measurement volumes. The shape of the vessel has the property that the ratio of the surface area of contents in the vessel to the volume of those contents is constant across the operational range. For any source of error in the estimation of vertical height of contents, the constant ratio of surface area to volume prevents the error from having a larger negative impact on overall measurement error as the volume being measured decreases. In other words, the vessel is just at good at measuring small volumes of contents as large volumes of contents.

Volumetric measurement device

Systems and techniques are disclosed for generating weighted machine learned models using multi-shard combiners. A learner in a machine learning system may receive labeled positive and negative examples and workers within the learner may be configured to receive either positive or negative examples. A positive and negative statistic may be calculated for a given feature and may either be applied separately in a model or may be combined to generate an overall statistic.

Using specialized workers to improve performance in machine learning

In one embodiment, a user of a social networking system requests to search for a place near the user's current location. The social networking system generates a list of places near the user's current location, select a sub-set from the list of places based on visibility and activity of the user and the user's social contacts for each place in the list, and returns the sub-set to the user.

Dynamic place visibility in geo-social networking system

A server system includes a processor and a data structure having an entry for a resource, the entry including a first sequence number. The server has communication procedures for receiving a request from a client to access the resource, where the request includes a second sequence number obtained from a service, and a resource request handling program. Upon receiving the request, the resource request handling program determines whether the server has any record of having previously received a request to access the resource. If not, the server returns a provisional rejection to the client, requiring the client to verify that it holds a lock on the specified resource. A provisional bit in the entry is initially set to indicate that the resource has not been accessed since the system was last initialized. The provisional bit is reset when a request to access the resource is granted.

Growing community of inventors

Mountain View, CA, United States of America

Joshua A Redstone