Optimization of Residual Gas Analysis mass spectrum through machine learning

Publié le : 8 octobre 2019

Context And background :
LCFC Lab offers a complete RGA testing solution for outgassing and hermetic studies of microelectronic and MEMS package cavities. Two ultra high vacuum apparatus has been developed as well as a devoted software package for qualitative and quantitative RGA data analysis. To strengthen the RGA activity, software optimization and improvement is required.

Job description:
Up to now, data extraction and analysis is based on physic fluid dynamic principles, and realized essentially through simple macro programming. As soon as the amount of data increase, the current analysis programs show limitations in term of computational efficiency and results reliability. Since RGA activity is growing up, it is mandatory to improve both the RGA analysis capability and the results reliability. The first task of the internship will be to rewrite these programs under the more appropriate Scilab or Matlab programming environment. The second part will be dedicated to the implementation of machine learning based strategies to facilitate the RGA mass spectrum interpretation. Our own database as well as externals free data base (for example provided by NIST) will be used to train and validate the neural network setup. Background with deep learning, software development, databases, and physics engineering skills is considered helpful.

If you are interested by the internship, please send your CV and a motivation letter to helene.duchemin@cea.fr

En naviguant sur notre site, vous acceptez que des cookies soient utilisés pour vous proposer des contenus et services adaptés à vos centres d’intérêts. En savoir plus