Abstract: Embodiments may include machine learning, including decision tree machine learning. Predictive variables may be selected for the machine learning through an iterative process. Predictive power of the predictive variables and collinearity between the predictive variables may be considered when selecting a set of the predictive variables for the machine learning.

Abstract: A system for database management and presentation processing of a graphical user interface includes a network interface interfacing with multiple data sources comprising internal systems and external systems, one or more processors, and an application service integration and communication layer to receive data including a live data feed. A data storage device stores schemas of data structures, and the data structures may include tables based on types and fields. The application service integration and communication layer normalizes and formats the live data feed according to the predetermined schemas, and stores the normalized and formatted procurement data in the database. A graphical user interface of true conditions assigned to procurement data in the live data feed is displayed.

Abstract: A machine learning based procurement system comprises a machine learning classifier to classify bids. The procurement system determines a price risk score and a supplier risk score for each of the bids based on the classifications, and determines if any of the bids are associated with a high-risk procurement based on comparing the price risk score and the supplier risk score to the respective threshold. The procurement system includes a graphical user interface that can display bid evaluation links, which are accessible to provide information explaining high-risk procurements.

Abstract: A machine learning based procurement system comprises a machine learning classifier to classify bids. The procurement system determines a price risk score and a supplier risk score for each of the bids based on the classifications, and determines if any of the bids are associated with a high-risk procurement based on comparing the price risk score and the supplier risk score to the respective threshold. The procurement system includes a graphical user interface that can display bid evaluation links, which are accessible to provide information explaining high-risk procurements.

Abstract: A machine learning classifier based procurement system determines a price risk score, a supplier risk score, and an item risk score for bids based on classifications performed by a machine learning classifier. The scores are compared to respective thresholds to determine if any of the bids are associated with a high-risk procurement.

Abstract: A machine learning classifier system includes a data set processing subsystem to generate a training set and a validation set from multiple data sources. Classifier hardware induces a classifier according to the training set, and tests the classifier according to the validation set. A buffer connected to the classifier hardware stores data objects to be classified, and a register connected to the classifier hardware stores outputs of the classifier, including classified data objects.

Abstract: A machine learning classifier system includes a data set processing subsystem to generate a training set and a validation set from multiple data sources. Classifier hardware induces a classifier according to the training set, and tests the classifier according to the validation set. A buffer connected to the classifier hardware stores data objects to be classified, and a register connected to the classifier hardware stores outputs of the classifier, including classified data objects.

Abstract: A machine learning classifier system includes a data set processing subsystem to generate a training set and a validation set from multiple data sources. Classifier hardware induces a classifier according to the training set, and tests the classifier according to the validation set. A buffer connected to the classifier hardware stores data objects to be classified, and a register connected to the classifier hardware stores outputs of the classifier, including classified data objects.

Abstract: Embodiments may include machine learning, including decision tree machine learning. Predictive variables may be selected for the machine learning through an iterative process. Predictive power of the predictive variables and collinearity between the predictive variables may be considered when selecting a set of the predictive variables for the machine learning.

Abstract: A system for database management and presentation processing of a graphical user interface includes a network interface interfacing with multiple data sources comprising internal systems and external systems, one or more processors, and an application service integration and communication layer to receive data including a live data feed. A data storage device stores schemas of data structures, and the data structures may include tables based on types and fields. The application service integration and communication layer normalizes and formats the live data feed according to the predetermined schemas, and stores the normalized and formatted procurement data in the database. A graphical user interface of true conditions assigned to procurement data in the live data feed is displayed.

Abstract: A paneled partition assembly is disclosed that can be used to selectively separate a first space from a second space. The paneled partition assembly can cover an opening when in a closed and extended position and uncover the opening when in an opened and retracted position. The paneled partition assembly can include a first panel and a second panel configured such that rotation of a handle on the first panel causes linear movement of the second panel. The partition assembly can include a pair of mating assemblies that selectively move from open and retracted positions on opposite sides of an opening to closed and extended positions in which opposite leading edges abut one another.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for selecting a forecastability strategy from among a group of forecastability strategies based on business outcomes. A group of predefined forecastability strategies is received along with historical supply chain management data. Each forecastability strategy represents a set of rules by which to determine whether each item of a plurality of items within a supply chain management system is to be managed using statistical forecasting. The historical supply chain management data represents past events in a supply chain associated with at least two items selected from the plurality of items within the supply chain management system. Each of the forecastability strategies is applied to the historical data in order to generate business outcomes for each of the forecastability strategies.

Abstract: A procurement analysis system may include a network interface to interface with one or more data sources over a network to receive procurement data and a data storage device to store model building data sets determined from the procurement data. A model generator module may determine procurement scoring models from the model building data sets, and a procurement risk analysis module may determine an evaluation capacity to review bids for a procurement, and evaluate bids for the procurement based on the procurement scoring models and the evaluation capacity to identify high-risk procurements associated with the evaluated bids.

Abstract: A method, computer program product, and apparatus comprise providing a Java developer kit class library comprising first Java classes; providing a Java Open Foundation class library comprising second Java classes each containing one or more of the first Java classes; providing a .NET SDK class library comprising .NET classes; providing a .NET language Open Foundation class library comprising first .NET language classes each containing one or more of the .NET classes; wherein each of the .NET language classes presents the same method signature as one of the second Java classes; receiving Java source code defining third Java classes, first calls to the first Java classes, and second calls to the second Java classes; translating the third Java classes into second .NET language classes; and translating the first calls to calls into the .NET SDK class library without modifying the second calls.

Abstract: A technique for post-processing a conventionally completed semiconductor device having a final passivation layer and bond pads exposed through the final passivation layer. The technique includes forming a protective film over the final passivation layer and exposed bond pads of the semiconductor device, and thereafter performing post-processing of the completed semiconductor device. Post-process structures, such as charge-coupled devices, can be formed above the protective film during this post-processing. Subsequent to the post-processing, the protective film is selectively etched to again expose the bond pads.

Abstract: A technique for post-processing a conventionally completed semiconductor device having a final passivation layer and bond pads exposed through the final passivation layer. The technique includes forming a protective film over the final passivation layer and exposed bond pads of the semiconductor device, and thereafter performing post-processing of the completed semiconductor device. Post-process structures, such as charge-coupled devices, can be formed above the protective film during this post-processing. Subsequent to the post-processing, the protective film is selectively etched to again expose the bond pads.

Abstract: A method and structure for reducing short circuits in semiconductor devices is disclosed. A three layer interlevel dielectric structure is formed over a semiconductor substrate, which typically comprises a first metallization level, M1. The three layer dielectric includes a first insulator layer, a middle spin-on glass (SOG) layer, and a top second insulator layer. The spin-on glass fills defects in the surface of the first insulator layer created during planarization using chemical-mechanical-polishing (CMP). Prior to deposition of the second insulator, a first via is etched through the SOG film and the first insulator layer to expose a portion of the semiconductor substrate, typically a conductive metal. A conductive metal is deposited into the first via and planarized to form a metal interconnection stud. Because the surface defects are filled and covered with the SOG film, none of the deposited metal enters the defects, and short circuits with the stud are greatly reduced.

Abstract: A method and structure for reducing short circuits in semiconductor devices is disclosed. A three layer interlevel dielectric structure is formed over a semiconductor substrate, which typically comprises a first metallization level, M1. The three layer dielectric includes a first insulator layer, a middle spin-on glass (SOG) layer, and a top second insulator layer. The spin-on glass fills defects in the surface of the first insulator layer created during planarization using chemical-mechanical-polishing (CMP). Prior to deposition of the second insulator, a first via is etched through the SOG film and the first insulator layer to expose a portion of the semiconductor substrate, typically a conductive metal. A conductive metal is deposited into the first via and planarized to form a metal interconnection stud. Because the surface defects are filled and covered with the SOG film, none of the deposited metal enters the defects, and short circuits with the stud are greatly reduced.