CER SEMINAR

Application of Pattern Recognition to Spectral Data of High and Low Temperature Plasma

Mehmet Fatih Yılmaz
May 16, 2018, 11:00am - 12:00pm, SERF Room 383

ABSTRACT:

The pattern recognition techniques such as principal component, linear discriminant analysis and others can be used as an alternative method to line intensity ratio diagnostics of the high energy density plasmas. Since these techniques can perform the spectral data in vectorial manner, they can easily indicate the polarization of the lines caused by the magnetic fields or electron beams. These techniques also reveals the hidden collective structures of high energy density plasmas such as Dirac cone like and Langmuir turbulence like structures especially in the presence of electron beams. Our ongoing research with the application of these techniques to the spectra of low temperature plasmas reveals the signatures of plasma condensation structures and plasma oscillations.

BIO:

Mehmet Fatih Yilmaz received his PhD in atomic spectroscopy of high energy density Z-pinch and laser-produced plasmas from University of Nevada, Reno in 2009. From 2010 to 2013, he worked as faculty in the Engineering Department of Mevlana University. In 2013, he was promoted to be chief research scientist at the Space Institute of Tubitak, Ankara. Since 2017, he has been serving as consultant to Engineering Department of University of Dammam, KSA for the spectroscopic investigations of low temperature plasmas and light-nanoparticles interactions. His current research is focusing on the applications of pattern recognition and machine learning techniques of X-pinch produced high energy density, laser and discharge-produced low temperature plasmas with the collaboration of Ecole Poly Technique, France and Dammam University, KSA.