Tensor decompositions are the “Swiss knife” of data science, data mining, machine learning and signal processing. They can be used for a variety of problems such as factor analysis; co-clustering; outlier/anomaly detection; and missing value estimation and interpolation; link prediction in time-evolving networks; multivariate time series forecasting; feature extraction in classification, as well as compression/de-noising and pattern recognition in signal processing. The recent research shows that tensor-based methods beat almost all non-tensor-based methods in a wide range of real-world applications. If we want to name a single tool, equivalent to deep learning but in the unsupervised setting that tool is perhaps tensor decompositions. Even, recently it is shown that tensor decompositions have applications in compressing the parameter space of deep learning models. The main reason behind this versatility is that the natural structure of many real-world datasets is highly multi-way in many applications and tensor decompositions are capable of capturing those complex interactions across various ways of data and thus provide a better and more natural model.
The Special session on “Tensor Analytics for Emerging Applications” aims at bringing together computer scientists, data scientists, and domain experts from various application areas to discuss the recent advances in algorithms, models, scalable solutions, and real-life applications.
We expect diverse submissions spanning and integrating fields such as healthcare (e.g., analysis of COVID-19 data and phenotyping), industry (e.g., IoT and sensors), internet platforms (e.g., recommender systems and time-evolving social network analysis), transportation (e.g. interpolation of traffic data and analysis of spatio-temporal data), neuroscience (e.g., modelling EEG and fMRI imaging data), and remote sensing (e.g., modelling hyperspectral images).
Paper Submission: June 6, 2021
Paper Notification: July 25, 2021
Paper Camera Ready Due: August 8, 2021
We invite submissions that report progress in either theoretical, technical or application aspects of Tensor Analytics. The topics include, but are not limited to the following:
Program Committee Members
Panos Markopoulos, Assistant Professor, Rochester Institute of Technology, USA
Kijung Shin, Assistant Professor, KIST, South Korea
Xiao Fu, Assistant Professor, Oregon State University, USA
Shaden Smith, Senior Research SDE, Microsoft, USA
Joyce C Ho, Assistant Professor, Emory University, USA
Kejun Huang, Assistant Professor, University of Florida, USA
Evangelos E. Papalexakis, Assistant Professor, University of California Riverside, USA
Hadi Fanaee-T, Assistant Professor, Halmstad University, Sweden
Maryam Amoozegar, Assistant Professor, Kerman Graduate University of Technology, Iran
Sofia Fernandes, Postdoc, University of Aveiro, Portugal
Mehran Yazdi, Professor, Shiraz University, Iran
Mansoor Rezghi, Associate Professor, Tarbiat Modares University , Iran
Farnaz Sedighin, Research Scientist, Skoltech, Russia
Leila Sorkhabi, Azad University, Iran