collaborators
play

Collaborators Joint work with Sarah Dean, Aurelia Guy, Horia Mania, - PowerPoint PPT Presentation

Collaborators Joint work with Sarah Dean, Aurelia Guy, Horia Mania, Nikolai Matni, Max Simchowitz, and Stephen Tu. trustable, scalable, predictable <latexit


  1. Collaborators Joint work with Sarah Dean, Aurelia Guy, Horia Mania, Nikolai Matni, Max Simchowitz, and Stephen Tu.

  2. trustable, scalable, predictable

  3. <latexit sha1_base64="oTGOPnlC3lpbuJxkZHAlqk3gehs=">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</latexit> <latexit sha1_base64="oTGOPnlC3lpbuJxkZHAlqk3gehs=">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</latexit> <latexit sha1_base64="oTGOPnlC3lpbuJxkZHAlqk3gehs=">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</latexit> <latexit sha1_base64="oTGOPnlC3lpbuJxkZHAlqk3gehs=">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</latexit> <latexit sha1_base64="DOx/ktybitgjChwuZWtodyh8jiA=">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</latexit> <latexit sha1_base64="DOx/ktybitgjChwuZWtodyh8jiA=">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</latexit> <latexit sha1_base64="DOx/ktybitgjChwuZWtodyh8jiA=">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</latexit> <latexit sha1_base64="DOx/ktybitgjChwuZWtodyh8jiA=">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</latexit> <latexit sha1_base64="Vs+14vGXEYCWQa4/aBIirWhHyZg=">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</latexit> <latexit sha1_base64="Vs+14vGXEYCWQa4/aBIirWhHyZg=">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</latexit> <latexit sha1_base64="Vs+14vGXEYCWQa4/aBIirWhHyZg=">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</latexit> <latexit sha1_base64="Vs+14vGXEYCWQa4/aBIirWhHyZg=">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</latexit> Optimal control u x hP T i minimize t = 1 C t ( x t , u t ) E e e s.t. x t + 1 = f t ( x t , u t , e t ) x t u t = π t ( τ t ) C t is the cost. If you maximize, it’s called a reward. x t is the state, u t is the input, e t is a noise process f t is the state-transition function τ t = ( u 1 , . . . , u t − 1 , x 0 , . . . , x t ) is an observed trajectory is the policy. This is the optimization decision variable. π t ( τ t )

Recommend


More recommend