Comparison of Advanced Machine Learning Tools for Disruption Prediction and Disruption Studies

Machine learning tools have been used since a long time ago to study disruptions and to predict their occurrence. On the other hand, the challenges posed by the quality and quantities of the data available remain substantial. In this paper, methods to optimize the training data set and the potential of kernels-based advanced machine learning tools are explored and assessed. Various alternatives, ranging from appropriate selection of the weights to the inclusion of artificial points, are investigated to improve the quality of the training data set. Support vector machines (SVM), relevance vector machines (RVMs), and one-class SVM are compared. The relative performances of the different approaches are initially assessed using synthetic data. Then they are applied to a relatively large database of JET disruptions. It is shown that in terms of final results, the optimization of the training databases proved to be very productive. Further, the RVM algorithm performs well when it is trained on a small set of discharges compared to the traditional methods.

[1]  George Eastman House,et al.  Sparse Bayesian Learning and the Relevance Vector Machine , 2001 .

[2]  A. Murari,et al.  Implementation of the Disruption Predictor APODIS in JET’s Real-Time Network Using the MARTe Framework , 2012, IEEE Transactions on Nuclear Science.

[3]  Hsuan-Tien Lin,et al.  A note on Platt’s probabilistic outputs for support vector machines , 2007, Machine Learning.

[4]  Piergiorgio Sonato,et al.  A prediction tool for real-time application in the disruption protection system at JET , 2007 .

[5]  Wendell Horton,et al.  Neural network prediction of some classes of tokamak disruptions , 1996 .

[6]  J. Contributors,et al.  Survey of disruption causes at JET , 2011 .

[7]  Bernhard Schölkopf,et al.  Improving the Accuracy and Speed of Support Vector Machines , 1996, NIPS.

[8]  Francesco Carlo Morabito,et al.  Fuzzy-neural approaches to the prediction of disruptions in ASDEX Upgrade , 2001 .

[9]  Bernhard Schölkopf,et al.  Estimating the Support of a High-Dimensional Distribution , 2001, Neural Computation.

[10]  Vladimir N. Vapnik,et al.  The Nature of Statistical Learning Theory , 2000, Statistics for Engineering and Information Science.

[11]  Isabelle Guyon,et al.  An Introduction to Variable and Feature Selection , 2003, J. Mach. Learn. Res..

[12]  Michael E. Tipping.  Bayesian Inference: An Introduction to Principles and Practice in Machine Learning , 2003, Advanced Lectures on Machine Learning.

[13]  J. Manickam,et al.  Chapter 3: MHD stability, operational limits and disruptions , 2007 .

[14]  J. Contributors,et al.  Statistical analysis of disruptions in JET , 2009 .

[15]  M. Sugihara,et al.  Disruption scenarios, their mitigation and operation window in ITER , 2007 .

[16]  Andrea Murari,et al.  An advanced disruption predictor for JET tested in a simulated real-time environment , 2010 .

[17]  Jet Efda Contributors,et al.  A cross-tokamak neural network disruption predictor for the JET and ASDEX Upgrade tokamaks , 2005 .

[18]  Chih-Jen Lin,et al.  LIBSVM: A library for support vector machines , 2011, TIST.

[19]  F C Schuller,et al.  Disruptions in tokamaks , 1995 .

[20]  Alessandra Fanni,et al.  Disruption forecasting at JET using neural networks , 2002 .

[21]  Bernhard Schölkopf,et al.  New Support Vector Algorithms , 2000, Neural Computation.

[22]  K. F. Mast,et al.  On-line prediction and mitigation of disruptions in ASDEX Upgrade , 2002 .

[23]  J. A. Leuer,et al.  Tokamak disruption alarm based on a neural network model of the high- beta limit , 1997 .

[24]  R. Yoshino.  Neural-net disruption predictor in JT-60U , 2003 .

[25]  T. Tajima,et al.  Forecast of TEXT plasma disruptions using soft X rays as input signal in a neural network , 1999 .

[26]  A. Polevoi,et al.  Overview of physics basis for ITER , 2003 .