Yali Bian
I am a senior research scientist at Human & AI Systems Research Lab (HAR), Intel Labs. I am working on designing systems that enable effective collaborations between human and interactive deep learning models. I got my Ph.D. in Computer Science, at Virginia Tech, working with Dr. Chris North at InfoVis Lab. Before that, I got an M.S. degree in Computer Science at Zhejiang University.
Human-AI Collaboration Interactive ML, eXplainable AI, Visual Analytics,
Featured Projects
Professional Experience
-
Establishing an AI-assisted image annotation system to reduce annotation efforts and accelerate DL model building for object detection tasks;
Building an exploratory data analysis platform with interpretable DL to help help stakeholders for decision making process. -
Built the ML pipeline to standardize the ML development life cycle and fully utilize existing ML platforms.
Built a Feature Store (Marketplace) system to manage, monitor, and serve batch and streaming features. -
Akuna Capital
Software Development EngineerMay. 2019 – Mar. 2020Led a team in building trade surveillance and compliance application to monitor and analyze live trading behaviors from the message streaming platform, and optimized the system processing efficiency by 72%. -
Facebook
Machine Learning Engineer InternMay 2018 – Aug 2018Machine Learning Group, Facebook, Seattle.Building alumni groups recommendation model for Facebook users through Sparse Neural Networks (SparseNN) and Gradient Boosting Decision Tree(GBDT). -
NOKIA Bell Labs
Machine Learning Research InternJun 2017 – Aug 2017Supervisor: Chitra Phadke; Huseyin UzunaliogluBuilding a smarter visual analytics system through semantic interaction and data-driven model selection to reduce human’s efforts on gaining insights from big and complex data.
Selected Publications
-
Semantic Explanation of Interactive Dimensionality Reduction.IEEE VIS 2021
-
DeepVA: Bridging Cognition and Computation through Semantic Interaction and Deep Learning.Proceedings of the IEEE VIS Workshop MLUI 2019: Machine Learning from User Interactions for Visualization and Analytics. VIS’19.
-
Evaluating Semantic Interaction on Word Embeddings via SimulationEValuation of Interactive VisuAl Machine Learning systems, an IEEE VIS 2019 Workshop
-
CKGHV: A comprehensive knowledge graph for history visualization.Proceedings of the 14th ACM/IEEE-CS Joint Conference on Digital Libraries. IEEE Press, 2014
-
Discovering Treatment Pattern in Traditional Chinese Medicine Clinical Cases by Exploiting Supervised Topic Model and Domain Knowledge.Journal of Biomedical Informatics 58: 260-267 (2015)Concept over time: the combination of probabilistic topic model with wikipedia knowledge.Expert Systems with Applications, Volume 60, 27-38 (2016)
Talks & News
-
Introduction to Data VisualizationThis session will cover the basics of matplotlib and the booked plotting library for interactive visualization. A basic knowledge of python is helpful but not necessary. Hands-on introduction to Data Visualization with Plot.ly. Knowledge of Python will be helpful as we will be using the Python API for the examples.
-
Introduction to Python for Scientific ComputingThis session will cover the basics of matplotlib and the booked plotting library for interactive visualization. A basic knowledge of python is helpful but not necessary. Hands-on introduction to Data Visualization with Plot.ly. Knowledge of Python will be helpful as we will be using the Python API for the examples.
-
D3-Workshop