Publications

Kang, Qiao, Jesper Larsson Träff, Reda Al-Bahrani, Ankit Agrawal, Alok Nidhi Choudhary, and Wei-Keng Liao. "Scalable Algorithms for MPI Intergroup Allgather and Allgatherv." Parallel Computing 85 (2019): 220-230.

Kang, Qiao, Ankit Agrawal, Alok Choudhary, and Wei-keng Liao. "Optimal Algorithms for Half-Duplex Inter-Group All-to-All Broadcast on Fully Connected and Ring Topologies."

Kang, Qiao, Jesper Larsson Träff, Reda Al-Bahrani, Ankit Agrawal, Alok Choudhary, and Wei-keng Liao. "Full-duplex inter-group all-to-all broadcast algorithms with optimal bandwidth." In Proceedings of the 25th European MPI Users’ Group Meeting, p. 1. 2018.

Kettimuthu, Rajkumar, Zhengchun Liu, Ian T. Foster, Peter H. Beckman, Alex Sim, Kesheng Wu, Wei-keng Liao, Qiao Kang, Ankit Agrawal, and Alok N. Choudhary. "Towards Autonomic Science Infrastructure: Architecture, Limitations, and Open Issues." In AI-Science@ HPDC, pp. 2-1. 2018.

Kang, Qiao, Wei-Keng Liao, Ankit Agrawal, and Alok Choudhary. "A Hybrid Training Algorithm for Recurrent Neural Network Using Particle Swarm Optimization-Based Preprocessing and Temporal Error Aggregation." In 2017 IEEE International Conference on Data Mining Workshops (ICDMW), pp. 812-817. IEEE, 2017.

Kang, Qiao, Wei-keng Liao, Ankit Agrawal, and Alok Choudhary. "A Filtering-based Clustering Algorithm for Improving Spatio-temporal Kriging Interpolation Accuracy." In Proceedings of the 25th ACM International on Conference on Information and Knowledge Management, pp. 2209-2214. ACM, 2016.

Projects

Optimizing MPI Intergroup Collective Communication Algorithms

This project designs and implement new algorithms for MPI collective communications with intergroup communicators. The results can be used to improve data communication performance among workflow components.

Architecture and Management for Autonomic Science Ecosystems

This project includes anomaly detection for supercomputers and data synchronization algorithms.

A Hybrid Training Algorithm for Recurrent Neural Network Using Particle Swarm Optimization-based Preprocessing and Temporal Error Aggregation

Designing, implementing and evaluating a new training algorithm for recurrent neural network aiming for faster convergence and lower error with exploitation and exploration. Synchronous PSO and enhanced back propagation through time are proposed to reduce error and improve convergence speed.

A Filtering-based Clustering Algorithm for Improving Spatio-temporal Kriging Interpolation Accuracy

Designing, implementing and evaluating a novel clustering algorithm that improves Kriging interpolation accuracy using two stages: filtering and reinforcing.