Abstract: Input on a plurality of attributes of a computing environment is provided to a machine learning module to produce an output value that comprises a risk score that indicates a likelihood of a potential malfunctioning occurring within the computing environment. A determination is made as to whether the risk score exceeds a predetermined threshold. In response to determining that the risk score exceeds a predetermined threshold, an indication is transmitted to indicate that potential malfunctioning is likely to occur within the computing environment. A modification is made to the computing environment to prevent the potential malfunctioning from occurring.

Abstract: A machine learning module is trained by receiving inputs comprising attributes of a computing environment, where the attributes affect a likelihood of failure in the computing environment. In response to an event occurring in the computing environment, a risk score that indicates a predicted likelihood of failure in the computing environment is generated via forward propagation through a plurality of layers of the machine learning module. A margin of error is calculated based on comparing the generated risk score to an expected risk score, where the expected risk score indicates an expected likelihood of failure in the computing environment corresponding to the event. An adjustment is made of weights of links that interconnect nodes of the plurality of layers via back propagation to reduce the margin of error, to improve the predicted likelihood of failure in the computing environment.

Abstract: Input on a plurality of attributes of a computing environment is provided to a machine learning module to produce an output value that comprises a risk score that indicates a likelihood of a potential malfunctioning occurring within the computing environment. A determination is made as to whether the risk score exceeds a predetermined threshold. In response to determining that the risk score exceeds a predetermined threshold, an indication is transmitted to indicate that potential malfunctioning is likely to occur within the computing environment. A modification is made to the computing environment to prevent the potential malfunctioning from occurring.

Abstract: A machine learning module is trained by receiving inputs comprising attributes of a computing environment, where the attributes affect a likelihood of failure in the computing environment. In response to an event occurring in the computing environment, a risk score that indicates a predicted likelihood of failure in the computing environment is generated via forward propagation through a plurality of layers of the machine learning module. A margin of error is calculated based on comparing the generated risk score to an expected risk score, where the expected risk score indicates an expected likelihood of failure in the computing environment corresponding to the event. An adjustment is made of weights of links that interconnect nodes of the plurality of layers via back propagation to reduce the margin of error, to improve the predicted likelihood of failure in the computing environment.

Abstract: A repository of key-value data may store a first object value having an internal structure of a hierarchy of sub-objects. The repository may receive a request to modify the first object, expressed as a function, without side effects, of one or more parameters. Values may be assigned to the one or more parameters by mapping from the parameters to locations in the hierarchy of sub-objects. A second object may be formed based on the return value of the function. The first object may be replaced in the repository by the second object.

Abstract: A machine learning module is trained by receiving inputs comprising attributes of a computing environment, where the attributes affect a likelihood of failure in the computing environment. In response to an event occurring in the computing environment, a risk score that indicates a predicted likelihood of failure in the computing environment is generated via forward propagation through a plurality of layers of the machine learning module. A margin of error is calculated based on comparing the generated risk score to an expected risk score, where the expected risk score indicates an expected likelihood of failure in the computing environment corresponding to the event. An adjustment is made of weights of links that interconnect nodes of the plurality of layers via back propagation to reduce the margin of error, to improve the predicted likelihood of failure in the computing environment.

Abstract: Input on a plurality of attributes of a computing environment is provided to a machine learning module to produce an output value that comprises a risk score that indicates a likelihood of a potential malfunctioning occurring within the computing environment. A determination is made as to whether the risk score exceeds a predetermined threshold. In response to determining that the risk score exceeds a predetermined threshold, an indication is transmitted to indicate that potential malfunctioning is likely to occur within the computing environment. A modification is made to the computing environment to prevent the potential malfunctioning from occurring.

Abstract: A repository of key-value data may store a first object value having an internal structure of a hierarchy of sub-objects. The repository may receive a request to modify the first object, expressed as a function, without side effects, of one or more parameters. Values may be assigned to the one or more parameters by mapping from the parameters to locations in the hierarchy of sub-objects. A second object may be formed based on the return value of the function. The first object may be replaced in the repository by the second object.

Abstract: One or more computing nodes located in a first region may maintain a first collection of data items. A second set of one or more computing nodes in a second region may maintain a collection of data items that is at least a partial replica of the first collection. Notifications of changes to the first collection may be transmitted, via a broadcast communications channel, to a replication module, which may be included in a client application. The replication module may transmit requests to update the second collection of data based on changes to the first collection. Conflicts may be resolved through a last-write wins policy.

Abstract: A client application may interact with a database service using a data interchange format with limited support for database data types or query representations. A map may be supported in the interchange format as a collection of name-value pairs. A database query may be parsed to form an abstract syntax tree. From the abstract syntax tree, a hierarchy of nested maps may be formed to represent the query, based on schemas mapping from regions of the abstract syntax tree to maps in the interchange format.

Abstract: A repository of key-value data may store a first object value having an internal structure of a hierarchy of sub-objects. The repository may receive a request to modify the first object, expressed as a projection of locations in the object to be updated and a function that, upon evaluation, returns values to be used to update the projected locations of the object. The repository may determine that the locations specified by the projections correspond to non-overlapping regions of the object and, based on the determination, update the object using the results of evaluating the function.

Abstract: A database may store an object comprising a hierarchy of sub-objects. Numeric identifiers may be assigned to names of the sub-objects and used in place of the names when writing the object to storage. Scalar values may be prefixed with length indicators. Collections may be prefixed with counts indicative of the number of elements in the corresponding collection. A portion of the object may be retrieved from storage by traversing a path from the root of the hierarchy to the desired portion. Length and count information may be used to skip over portions of the object during the traversal.

Abstract: A method is provided for protecting an on-line resource using a HIP challenge. The method includes receiving a request to access the on-line resource from a remote client. A HIP challenge is presented to a user associated with the remote client. If a successful response to the HIP challenge is received from the user, a previous response pattern of the user is compared to known response patterns of humans and machines. The user is allowed to access to the on-line resource if the comparison indicates that the user is likely a human.

Abstract: A method for generating one or more human interactive proofs (HIPs) is described herein. A HIP request may be received. One of a plurality of HIP engines may be selected using a randomization algorithm. Each HIP engine may have a distinct algorithm for generating the one or more HIPs. The one or more HIPs may be generated using the one of the plurality of HIP engines.

Abstract: A method is provided for protecting an on-line resource using a HIP challenge. The method includes receiving a request to access the on-line resource from a remote client. A HIP challenge is presented to a user associated with the remote client. If a successful response to the HIP challenge is received from the user, a previous response pattern of the user is compared to known response patterns of humans and machines. The user is allowed to access to the on-line resource if the comparison indicates that the user is likely a human.

Abstract: A method for generating one or more human interactive proofs (HIPs) is described herein. A HIP request may be received. One of a plurality of HIP engines may be selected using a randomization algorithm. Each HIP engine may have a distinct algorithm for generating the one or more HIPs. The one or more HIPs may be generated using the one of the plurality of HIP engines.

Abstract: An occupant classification system utilizes a rules-based expert system to automatically classify the occupant of a seat for the purposes of airbag deployment. The invention provides users with the ability to create, test, and modify the image attributes or “features” used by the expert system to classify occupants into one of several predefined occupant-type categories. Users are also provided the ability to create, test, and modify the processes utilizing those chosen features. The user of the invention designs the features and the algorithms used by the expert system classifier. A feature extractor is used to extract features from an image of the occupant and surrounding seat area, and the values relating to those features are sent in a vector of features to the expert system classifier. The expert system classifier classifies the image of the occupant according to the internal rules for that classifier.

Abstract: The method of the present invention involves a two-phase process for in-situ retorting and recovery of carbonaceous material contained within typical subterranean tar sand formations, and includes formation of conventional arrays of in-seam ducts, and positioning heating devices to heat a section of the formation over a large extent thereof. The operation of the heating devices in the first phase is controlled to provide heat into the formation without burning of the carbonaceous material therein, resulting in development of a quasi-stable zone of pyrolysis about the heating duct, to thermally crack the carbonaceous material producing various organic liquid oil fractions and derived condensible vapors and non-condensible gases. The products produced thereby are then withdrawn through a suitable array of collection wells.