Abstract: A distributed machine learning framework implemented with heterogeneous data platforms reduces data copying and exploits memory/computation resources of the different data platforms. A configuration component includes information to set up the system. A persistency component manages storage of data and a model trained by machine learning. A proxy repository includes predefined proxies for communication between heterogeneous data platform nodes and execution of the machine learning procedure. A machine learning execution component comprises three layers. A bottom work node layer within the data platform performs computations of the machine learning procedure. A middle server node layer comprising one server node per data platform, communicates with the work nodes to coordinate jobs on that data platform. An upper layer comprises a central server node communicating with server nodes and coordinating jobs of the different platforms.

Abstract: A rule handler may receive a first routing rule and a second routing rule, each routing rule characterizing at least one potential physical transportation route. A rule converter may convert the first routing rule and the second routing rule into a first finite automaton (FA) and a second finite automaton, respectively, and may combine the first finite automaton and the second finite automaton into a finite automaton network graph. A graph comparator may then compare the finite automaton network graph to existing physical transportation routes, and a route selector may select, based on the comparing, at least one feasible physical transportation route from the existing physical transportation routes, the at least one feasible physical transportation route being an instance of the at least one potential physical transportation route.

Abstract: The embodiments provide a system for matching orders with incoming shipments. The system may include a product shipment handler configured to receive product shipment information specifying incoming products, a product order handler configured to receive product order information specifying a plurality of product orders and input parameter information associated with each product order, a job dispatcher configured to create a plurality of jobs based on the product shipment information and the product order information, a modeling tool configured to model each job as a separate constrained linear optimization problem, and a linear solver configured to solve each constrained linear optimization problem based on the input parameter information and the product shipment information including generating a delivery plan for each product order within a respective job that matches at least one incoming product with a corresponding product order over a period of time.

Abstract: A system for self-adaptive user reading preference learning and screen layout optimization for multi-media information is disclosed. The system is particular useful for devices with small display form factors, such as mobile devices. Data flow diagram (DFD) technique is used to represent a model of the screen layout for visualization. The system processes input information, which includes user information, device information, context information, and the news. The delivery news is filtered in the news processing system and presented to a user based the user's interest, whereas device information and context information, are also factors that take effect in the modeling.

Abstract: A system and method for analyzing documents, such as posts, on-line reviews and comments from people based on topics of the documents, to determine general sentiment of users is disclosed. Topics from the documents and their corresponding sentiment polarities are extracted. The documents are regarded to be constituted by a series of topics. The sentiment for a topic is represented by a quadruple (k, so, h, i), where k is the topic, so is the sentiment opinion, h is the comment or post holder, and i is the document. A quintuple (k, sup, p, n, ne) is used to illustrate the topics and corresponding sentiments and is stored in S, where sup indicates the frequency of the topic, and p (positive), n (negative) and ne (neutral) are different types of opinions of the users. From the quintuple set S, every topic is related to three kinds of sentiment opinions (positive, negative, and neutral), enabling determination of popular topics in documents as well as the users' sentiment polarities.

Abstract: A distributed machine learning framework implemented with heterogeneous data platforms reduces data copying and exploits memory/computation resources of the different data platforms. A configuration component includes information to set up the system. A persistency component manages storage of data and a model trained by machine learning. A proxy repository includes predefined proxies for communication between heterogeneous data platform nodes and execution of the machine learning procedure. A machine learning execution component comprises three layers. A bottom work node layer within the data platform performs computations of the machine learning procedure. A middle server node layer comprising one server node per data platform, communicates with the work nodes to coordinate jobs on that data platform. An upper layer comprises a central server node communicating with server nodes and coordinating jobs of the different platforms.

Abstract: Techniques are described for forming a machine learning model vector, or just model vector, that represents a weighted combination of machine learning models, each associated with a corresponding feature set and parameterized by corresponding model parameters. A model vector generator generates such a model vector for executing automated machine learning with respect to historical data, including generating the model vector through an iterative selection of values for a feature vector, a weighted model vector, and a parameter vector that comprise the model vector. Accordingly, the various benefits of known and future machine learning algorithms are provided in a fast, effective, and efficient manner, which is highly adaptable to many different types of use cases.

Abstract: Techniques are provided for rendering network applications in a highly-customized manner, in which, for example, user interactions with one or more network applications using devices having different screen sizes are analyzed and used to assign user preferences and priorities with respect to the one or more network application(s). In this way, users may be provided with desired and useful content in a convenient manner, while application providers may have their content rendered in a manner that increases a likelihood of achieving an intended result (e.g., consummating a sale or other transaction, or eliciting some other desired reaction from the user).

Abstract: A system for forecast modeling includes at least one processor and at least one database that is operably coupled to the at least one processor. The database includes a time series data module that is configured to store time series data for a domain, an exogenous data module that is configured to store exogenous data associated with multiple exogenous factors and a feedback module that is configured to collect and store feedback data from multiple online users, where the feedback data is related to the exogenous data and the exogenous factors. The system includes a data pre-processor module that is configured to use the at least one processor to identify and select a portion of the exogenous factors using the feedback data collected from the online users for use in a forecast model in combination with the time series data for the domain.

Abstract: A system for self-adaptive user reading preference learning and screen layout optimization for multi-media information is disclosed. The system is particular useful for devices with small display form factors, such as mobile devices. Data flow diagram (DFD) technique is used to represent a model of the screen layout for visualization. The system processes input information, which includes user information, device information, context information, and the news. The delivery news is filtered in the news processing system and presented to a user based the user's interest, whereas device information and context information, are also factors that take effect in the modeling.

Abstract: A system and method for analyzing documents, such as posts, on-line reviews and comments from people based on topics of the documents, to determine general sentiment of users is disclosed. Topics from the documents and their corresponding sentiment polarities are extracted. The documents are regarded to be constituted by a series of topics. The sentiment for a topic is represented by a quadruple (k, so, h, i), where k is the topic, so is the sentiment opinion, h is the comment or post holder, and i is the document. A quintuple (k, sup, p, n, ne) is used to illustrate the topics and corresponding sentiments and is stored in S, where sup indicates the frequency of the topic, and p (positive), n (negative) and ne (neutral) are different types of opinions of the users. From the quintuple set S, every topic is related to three kinds of sentiment opinions (positive, negative, and neutral), enabling determination of popular topics in documents as well as the users' sentiment polarities.

Abstract: Techniques are described for forming a machine learning model vector, or just model vector, that represents a weighted combination of machine learning models, each associated with a corresponding feature set and parameterized by corresponding model parameters. A model vector generator generates such a model vector for executing automated machine learning with respect to historical data, including generating the model vector through an iterative selection of values for a feature vector, a weighted model vector, and a parameter vector that comprise the model vector. Accordingly, the various benefits of known and future machine learning algorithms are provided in a fast, effective, and efficient manner, which is highly adaptable to many different types of use cases.

Abstract: In accordance with aspects of the disclosure, systems and methods are provided for managing context aware recommendations by providing recommendations to a user in response to a query related to the user by integrating contextual information of a context related to the user in a recommendation model while considering a granular structure of the context and the contextual information thereof.

Abstract: Described herein is a technology for facilitating preference-based data representation. In accordance with one aspect of the technology, preference information is acquired from a user. Rank scores of objects are generated based at least in part on the user preference information. The objects are grouped into one or more clusters of objects based on the rank scores. A visualization of the one or more clusters of objects is then generated.

Abstract: A computer-implemented method for a booking based demand forecast problem includes converting time series data into a multivariate time series, training a multivariate time series model using the converted multivariate time series, forecasting results using the multivariate time series model and aggregating the results.

Abstract: A rule handler may receive a first routing rule and a second routing rule, each routing rule characterizing at least one potential physical transportation route. A rule converter may convert the first routing rule and the second routing rule into a first finite automaton (FA) and a second finite automaton, respectively, and may combine the first finite automaton and the second finite automaton into a finite automaton network graph. A graph comparator may then compare the finite automaton network graph to existing physical transportation routes, and a route selector may select, based on the comparing, at least one feasible physical transportation route from the existing physical transportation routes, the at least one feasible physical transportation route being an instance of the at least one potential physical transportation route.

Abstract: A computer-implemented method includes receiving, at a rental company, a customer's order to hire a transport equipment unit of a first type beginning on a start date. The rental company has an inventory of transport equipment units for hire and a network of diverse storage sites. The computer-implemented method further includes locating one or more transport equipment units of the first type in the rental company's inventory of transport equipment available for hire by the start date, computing the rental company's costs associated with picking up each of the located transport equipment units of the first type for delivery to the customer, and selecting, based on consideration of the rental company's picking up costs, one of the located transport equipment units for delivery to customer.

Abstract: A system for forecast modeling includes at least one processor and at least one database that is operably coupled to the at least one processor. The database includes a time series data module that is configured to store time series data for a domain, an exogenous data module that is configured to store exogenous data associated with multiple exogenous factors and a feedback module that is configured to collect and store feedback data from multiple online users, where the feedback data is related to the exogenous data and the exogenous factors. The system includes a data pre-processor module that is configured to use the at least one processor to identify and select a portion of the exogenous factors using the feedback data collected from the online users for use in a forecast model in combination with the time series data for the domain.

Abstract: Described herein is a technology for facilitating preference-based data representation. In accordance with one aspect of the technology, preference information is acquired from a user. Rank scores of objects are generated based at least in part on the user preference information. The objects are grouped into one or more clusters of objects based on the rank scores. A visualization of the one or more clusters of objects is then generated.

Abstract: In accordance with aspects of the disclosure, systems and methods are provided for managing context aware recommendations by providing recommendations to a user in response to a query related to the user by integrating contextual information of a context related to the user in a recommendation model while considering a granular structure of the context and the contextual information thereof.