Updated time: 2021-02-19 15:07:18
The DS2SHEAR screener can pick quickly and accurately the most important set of factors among a large number of factors that have significant impacts towards the response in an analysis. It will become one of the essential key techniques in many big data analytics system. It is a remedy that many analytics system need, no matter in the applications of artificial intelligence, industry, biomedicine or any other fields.
- DS2SHEAR screener combines the advantages of high detection accuracy and computational efficiency from the Dantzig Selector and the immuneion of the curse of dimensionality from the SRRS.
- The automatic factor coding method to transform a continuous factor dataset into a supersaturated design is new.
- The automatic selection on the correct underlying model (linear or logistic) for different response types is new.

Fig1. The flowchart of the DS2SHEAR screener for screening experiments.
06T-1091224
2021-02-19
KNOW-HOW
- This know-how is mainly used for detecting significant factors efficiently and stably among a large number of factors towards the responses in experimental analysis.
- It can be applied to classification problems in artificial intelligence and image processing, to detect key inputs in classification from the experimental data.
- It can be applied to diagnosis problems in biomedical science, to look for important biomarkers or phenotypes from biomedical data.
Frederick Kin Hung, Phoa、Jin-Wen Huang
If you are interested in this technology, please fill in the information on the webpage under the manufacturer option (Request for Technology Licensing, Material Transfer, or Sponsored Research) on the website of this office

Fig2. A Profile Plot of analyzing an order-of-addition experiment by DS2SHEAR Screener.

Dantzig-Selector-based 2-step Supersaturated and High-dimensional Experimental Analysis Recursive (DS2SHEAR) Screener