2 msc universit paris diderot umr cnrs 7057 paris
play

2 MSC, Universit Paris Diderot, UMR CNRS-7057, Paris 3 MAS, Centrale, - PowerPoint PPT Presentation

Adel Mezine 1 , Artmis Llamosi 2 , Vronique Letort 3 , Michle Sebag 4 , Florence dAlch -Buc 1,5 1 IBISC, Universit dEvry - Val dEssonne, Evry 2 MSC, Universit Paris Diderot, UMR CNRS-7057, Paris 3 MAS, Centrale,


  1. Adel Mezine 1 , Artémis Llamosi 2 , Véronique Letort 3 , Michèle Sebag 4 , Florence d’Alché -Buc 1,5 1 IBISC, Université d’Evry - Val d’Essonne, Evry 2 MSC, Université Paris Diderot, UMR CNRS-7057, Paris 3 MAS, Centrale, Châtenay-Malabry 4 LRI UMR CNRS-8623, INRIA Université Paris-Sud, Université Paris Saclay 5 LTCI UMR CNRS-5141, Telecom ParisTech, Université Paris Saclay

  2. Motivations Parameter and hidden state estimation Issues: Some parameters are non-identifiable in practice 1. 2. Perturbation experiments able to raise non-identifiability exist but are very expensive. 2/5

  3. EDEN: Experimental Design for Estimation in a Network Solution: Active learning that performs both system identification and sequential Design Of Experiments (DOE) under a fixed budget. Autonomous learning Parameter Estimation: The model is estimated from the current dataset using a global optimization procedure. Experimental Design : UCT -based active learning is applied to suggest the most promising sequence of experiments. ( UCT = MCTS with UCB - policy ) 3/5

  4. Dream7 Challenge Gene regulatory network Ordinary differential equations Budget = 10,000 credits Measurement Cost (credits) Perturbation Cost (credits) Mirco-array (50 points) 500 Knock-Out 800 Mirco-array (100 points) 1000 Knock-Down 350 Protein fluorescence (200 points) 400 Over-Expression 450 Gel-shift assay (two parameters) 1600 4/5

  5. Want to know more? Come see our poster! 5/5

  6. References � Browne, C. B., Powley, E., Whitehouse, D., Lucas , S. M., Cowling, P, I., Rohlfshagen, P., . . . & Colton, S.: A survey of Monte-Carlo tree search methods. Intelligence and AI (2012) � Kocsis, L., and Szepesvári, C.: Bandit based Monte-Carlo planning. ECML-06. (2006). � Llamosi, A., Mezine, A., Sebag M., Letort V., d’ Alché – Buc F. Experimental design in dynamical system identification: a bandit-based active learning approach . ECML/PKDD Proceedings, LNCS, Springer (2014). � Meyer, P., et al. Network topology and parameter estimation: from experimental design methods to gene regulatory network kinetics using a community based approach. BMC Systems Biology (2014). � Quach , M., Brunel, N., and d’ Alché – Buc, F.: Estimating parameters and hidden variables in non-linear state-space models based on odes for biological networks inference. Bioinformatics (2007). � Rolet, P., Sebag, M., and Teytaud, O.: Boosting active learning to optimality: A tractable Monte-Carlo, billiard-based algorithm. ECML/PKDD (2009). 6

Recommend


More recommend