International Business Machines Corporation

Jun Song Wang

Jianbin Tang

Jing James Xu

Qing Fa Wang

Ji Hui Yang

Yong Hua Lin

Zhen Bo Zhu

Rong Yan

Yonghua Lin

Chao Zhu

Hai Zhan

Dong Hai Yu

Qi Jun Wang

Xiao Ming Ma

Si Er Han

Xiaoyang Yang

Xin Hao Cheng

Song Bo

Josephine B Chang

Hendrik F Hamann

Theodore Gerard Van Kessel

Ramachandran Muralidhar

Yu Meng Li

Chuang Gan

Bei Chen

Hang Liu

Dakuo Wang

Chao Xue

A Peng Zhang

Dhavalkumar C Patel

Long H Vu

Syed Yousaf Shah

Mingqiang Li

Yang Zhang

Yao Dong Liu

Wen Pei Yu

Jiang Bo Kang

John Dillon Eversman

Voranouth Supadulya

Jing Min Xu

Jian Jun Wang

Kai Lin Li

Xue Ying Zhang

Thanh Lam Hoang

Lei Gao

Peter Daniel Kirchner

Xing Wei

Horst Cornelius Samulowitz

Chun Lei Xu

Craig Reginald Tomlyn

Saket Sathe

Gregory Bramble

Xuan-Hong Dang

Xinhao Cheng

Chong Liu

Rui Wang

Abel Nicolas Valente

Arunima Chaudhary

Yan Zhao Liu

Jin Rong Wang

Daniel Karl I Weidele

Lin Ju

Bo Song

Yang Y Yang

Jia Xing Tang

Xian Wei Yi

Jishuo Yang

Described are techniques for machine learning library management. The techniques include generating a table including a plurality of machine learning libraries and their current versions that are used in a deployed machine learning platform (MLP) instance, a first available version upgrade for a first machine learning library of the plurality of machine learning libraries, a security indication associated with the first available version upgrade relative to a current version implemented by the first machine learning library, and a compatibility indication between the first available version upgrade and the current version of the first machine learning library. The techniques further include generating a recommendation related to upgrading the first machine learning library based on the security indication and the compatibility indication.

Compatible and secure software upgrades

Provided are a computer-implemented method, a system, and a computer program product. The method comprises extracting features from a plurality of base models in an ensemble model. The plurality of base models are configured to provide respective prediction results. The ensemble model is configured to provide an overall prediction result from the prediction results of the plurality of base models. The features are associated with time performance of the base models. The method further comprises clustering the plurality of base models into a plurality of clusters based on the extracted features. The method further comprises assigning the plurality of base models to a plurality of parallel computation units based on the plurality of clusters.

Parallelized scoring for ensemble model

In an approach for detecting web browsing subject-oriented event interactions and intelligently organizing web pages based on insights from important interactions for better exploration and efficient management, a processor extracts time series data associated with a plurality of web browsing events based on browsing historical actions of a user. A processor identifies the subject of each web browsing event. A processor determines major events based on the time series data and subjects of the plurality of web browsing events. A processor organizes the plurality of web browsing events based on subject hierarchy and timeline from the time series data. A processor highlights one or more uniform resource locators based on the subject hierarchy and timeline.

Web smart exploration and management in browser

Systems and techniques for determining an idle timeout for a cloud computing session are described. An example technique includes determining a first one or more attributes associated with a user of the cloud computing session and determining a second one or more attributes associated with an operation of the cloud computing session. An idle timeout for the cloud computing session is determined, based at least in part on the first one or more attributes and the second one or more attributes. User activity is monitored during the cloud computing session. Upon determining, based on the monitoring, an absence of the activity of the user within a duration of the idle timeout, the cloud computing session is terminated.

Adjusting idle timeout for a session

Systems, computer-implemented methods, and computer program products to facilitate visualization of a model selection process are provided. According to an embodiment, a system can comprise a memory that stores computer executable components and a processor that executes the computer executable components stored in the memory. The computer executable components can comprise an interaction backend handler component that obtains one or more assessment metrics of a model pipeline candidate. The computer executable components can further comprise a visualization render component that renders a progress visualization of the model pipeline candidate based on the one or more assessment metrics.

Visualization of a model selection process in an automated model selection system

Techniques regarding one or more automated machine learning processes that analyze time series data are provided. For example, one or more embodiments described herein can comprise a system, which can comprise a memory that can store computer executable components. The system can also comprise a processor, operably coupled to the memory, and that can execute the computer executable components stored in the memory. The computer executable components can comprise a time series analysis component that selects a machine learning pipeline for meta transfer learning on time series data by sequentially allocating subsets of training data from the time series data amongst a plurality of machine learning pipeline candidates.

Automated machine learning pipeline generation

Methods, computer program products, and/or systems are provided that perform the following operations: obtaining a series of indicator diagrams corresponding to strokes of a pumpjack over a specific time duration, dividing each indicator diagram into a plurality of location segments in a direction of location of the rod; obtaining load difference features between upstroke loads and corresponding downstroke loads in the plurality of location segments; identifying a location segment with an abnormal load difference feature based on a time series data of load difference feature corresponding to one of the plurality of location segments, the time series data of load difference feature including a series of data points of load difference feature of the one of the plurality of location segments in time order; and providing an indication of a potential problem based, at least in part, on the identification of the location segment with an abnormal load difference feature.

Identifying potential problems in a pumpjack

A method for instantiating a first specific model, as one record, based on a model scoring template, clustering a plurality of models into a plurality of groups to obtain a plurality of clusters, computing a model contribution weight value for the given cluster, choosing a set of top model(s) from the plurality of models, and updating the model(s) of the set of top model(s) dynamically.

Effective ensemble model prediction system

Embodiments relate to a system, computer program product, and method for determining missing values in respective data records with an explanatory analysis to provide a context of the determined values. Such method includes receiving a dataset including incomplete data records that are missing predictors and complete data records. A model is trained with the complete data records and candidate predictors for the missing predictors are generated. A predictor importance value is generated for each candidate predictor and the candidate predictors that have a predictor importance value in excess of a first threshold value are promoted. Respective promoted candidate predictors are inserted into the respective incomplete data records, thereby creating tentative data records. The tentative data records are injected into the model, a fit value is determined for each of the tentative data records, and a tentative data record with a fit value exceeding a second threshold value is selected.

Explanative analysis for records with missing values

Embodiments generally relate to determining traffic congestion patterns. In some embodiments, a method includes identifying congestion events for each road of a plurality of roads in a road network, where each congestion event indicates a drop in average vehicle speed below a predetermined speed threshold for a particular road in the road network, and where the congestion events span a predetermined time period. The method further includes determining local clusters of the congestion events based on one or more road condition parameters, where each local cluster defines a local congestion pattern for a particular road of the plurality of roads in the road network. The method further includes grouping the local clusters into one or more global clusters based on the one or more road condition parameters, where the global clusters define global congestion patterns in the road network.

Determining traffic congestion patterns

Embodiments for managing feature engineering with relational data are provided. A graphical user interface (GUI) that provides a user with the ability to upload a plurality of tables, select joins between the plurality of tables, and select keys for the joins is provided. Responsive to receiving user input indicative of selecting joins between the plurality of tables and selecting keys for the joins utilizing the GUI, the user selections are automatically validated and actions associated with at least some of the plurality of tables are dynamically performed based on the user selections. Information associated with the user selections and the validating is provided. The information includes a recommendation to link a third key in the at least some of the plurality of tables to a fourth key in the at least some of the plurality of tables.

Methods and systems for visual management of relational data

A method, computer system, and a computer program product for detecting a plurality of influential factors for traffic congestion is provided. The present invention may include identifying one or more influential roads and one or more influential routings associated with an object road in a target road network. The present invention may also include quantifying an impact associated with the identified one or more influential roads and the identified one or more influential routings associated with the object road, wherein at least one Impact Matrix is utilized to determine a road-by-road influence in connection to the target road network.

Detecting influential factors for traffic congestion

A method, system and computer program product are provided for detecting vehicle queue events and managing traffic flow. A computing system recognizes whether a queue event occurred for each vehicle located in an area of interest based on collected vehicle data. The area of interest includes an intersection, and the vehicle data for the each vehicle includes location information and speed information. The location information further includes a distance to an intersection. The computing system identifies differences in queue length among queues in the area of interest based on the vehicle data and determines queue indicators for each of the queues in the area of interest. Based on queue indicators for each of the queues in in the area of interest generated over multiple sampling periods, traffic signal lights at the intersection in the area of interest are managed.

Detection of vehicle queueing events on a road

Growing community of inventors

Beijing, China

Jun Song Wang