NTU’s end-to-end AI prediction method is very smart

Select materials with excellent 99% CCS performance from hundreds of thousands of MOFs in minutes

How to predict the properties of hundreds of thousands of materials in minutes to find carbon capture and storage (CCS) metal organic frameworks (MOFs) with superior performance? Recently, Lu Cunxing, a 2020 graduate student instructed by Associate Professor Wan Xili of the School of Computer Science and Technology of Nanjing University of Technology, developed an end-to-end artificial intelligence (AI) prediction method based on deep learning for the carbon capture performance of MOFs

"CO2 is the main greenhouse gas that causes global warming.

According to Wan Xili, the prediction method developed by Lu Cunxing has two major advantages

Wan Xili said that in recent years, although there have been studies devoted to solving this problem, there are some limitations.

"My computational approach avoids the time-consuming and labor-intensive drawbacks of existing computations, and develops an end-to-end prediction method that does not require constructing descriptors, only takes crystallographic information files (CIF) as input, and adaptively learns the effects The high-dimensional characteristics of the performance can be used to predict the performance of MOFs quickly and accurately

This calculation method with self-learning ability will evaluate the gap between the predicted value and the real value at the end of each cycle, and then adjust its own parameters to reduce the gap, and minimize the error after multiple cycles, so as to achieve a more accurate quasi-function