<< /S /GoTo /D (section.5.4) >> endobj << /S /GoTo /D (subsection.1.1.3) >> << /S /GoTo /D (section.1.1) >> 77 0 obj endobj 136 0 obj
endobj endobj 21 0 obj (Exercice 2.10) 1.3 Different kinds of Monte Carlo simula-tions There are at least three different kinds of Monte Carlo simulations: • Transport simulations. 101 0 obj (Convergence vers la loi stationnaire.) endobj (G\351n\351rer des variables al\351atoires) (Le g\351n\351rateur de Mersenne Twister) (D\351termination de la p\351riode optimale) endobj << /S /GoTo /D (section.8.1) >> endobj %PDF-1.5 (Autres types de convergence) endobj endobj
2. endobj
<< /S /GoTo /D (subsection.8.1.2) >> endobj Monte Carlo, simulation, MCMC, estimation, optimization Abstract Many quantitative problems in science, engineering, and economics are nowadays solved via statistical sampling on a computer. endobj (Test d'uniformit\351) << /S /GoTo /D (section.4.1) >> This example differs in at least the two following ways from usual MC simulations: ... function (pdf), p(x). endobj << /S /GoTo /D (subsection.7.0.1) >>
endobj 48 0 obj a neutron) that reaches a shield.
<< /S /GoTo /D (chapter.2) >> 100 0 obj Cette méthode est intéressante puisqu'elle peut s'appliquer à des problèmes de grande dimension, comme par exemple pour calculer l'espérance de rendement sur l'ensemble d'un marché boursier. (Th\351or\350me ergodique) /Length 315 29 0 obj 65 0 obj 137 0 obj (exercices 2.07) 5 0 obj
endobj
121 0 obj 116 0 obj (Du bon choix des param\350tres) 41 0 obj (Simulation d'une loi B\352ta via Metropolis Hasting) << /S /GoTo /D (section.5.1) >> (Vitesse de convergence et temps de calcul) endobj
endobj endobj endobj (Test de Gelman Rubin) << /S /GoTo /D [186 0 R /Fit] >> endobj
186 0 obj << /Height 215 endobj
120 0 obj 33 0 obj endobj (III Simulation de Cha\356nes de Markov et algorithme de Metropolis Hastings) endobj endobj endobj << /S /GoTo /D (chapter.8) >> endobj 53 0 obj 25 0 obj endobj 85 0 obj endobj endobj 132 0 obj endobj 153 0 obj endobj
endobj << /S /GoTo /D (part.4) >>
/MediaBox [0 0 595.276 841.89] 185 0 obj endobj x��ut[W>��[��ޛ:�r�ә��N��Ift��33�efff[f�e�,[`K�l������7I�6�dS�k���V������zf�fV��7�'W�GVFGW��W�uxtevnm|j���Z���w�{��qMϬ�N������!|�68���9[�< m'?������c�}�}��s�ʣ�?8�<>�dy����������������56�6�>��m�mjjmiqknncWz2���d֤�fD�G�����؈����OOWwc����¢R_o��Ԭ���&�//o��ۧm�� b��f��s�T�疆�M�% 13 0 obj
133 0 obj (G\351n\351rer de l'al\351atoire: une n\351cessit\351 en simulation ) most MC methods are general and flexible enough to work with models of arbitrary complexity. << /S /GoTo /D (subsection.1.1.1) >>
169 0 obj
24 0 obj
endobj << /S /GoTo /D (section.8.6) >>
/Resources 190 0 R endobj endobj 108 0 obj 129 0 obj >> (M\351thode des variables antith\351tiques) 32 0 obj
stream 191 0 obj << (Cas de variables continues) (IV Exercices)
148 0 obj
(Loi conditionnelle) endobj endobj << /S /GoTo /D (section.1.2) >> 84 0 obj 172 0 obj 165 0 obj 92 0 obj endobj
Théorème central limite Soit (Xn)n‚1 une suite de variables aléatoires réelles i.i.d. endobj << /S /GoTo /D (chapter.6) >> endobj endobj For each point, determine whether it lies inside the unit circle, the red region in Figure ??.
endobj endobj endobj
endobj (Algorithme de Metropolis Hasting) << /S /GoTo /D (section.8.7) >> << /S /GoTo /D (subsection.2.1.1) >> (Test sur l'efficacit\351 des g\351n\351rateurs)
<< /S /GoTo /D (subsection.2.3.1) >>
61 0 obj 4 0 obj << /S /GoTo /D (subsection.1.2.2) >> 145 0 obj
(M\351thodes de R\351duction de la variance) (I G\351n\351rer de l'al\351atoire) << /S /GoTo /D (section.2.2) >>
128 0 obj endobj << /S /GoTo /D (section.8.4) >>
/Filter /FlateDecode 56 0 obj 45 0 obj << /S /GoTo /D (section.4.2) >>
<< /S /GoTo /D (chapter.5) >> endobj << /S /GoTo /D (chapter.4) >> << /S /GoTo /D (subsection.2.1.2) >> 109 0 obj endobj Introduction A brief overview Buffon’s experiment Monte Carlo simulation 1 Sample an u 1 ˘U[0;1) and u 2 U[0;1) 2 Calculate distance from a line: d = u 1 t 3 Calculate angle between needle’s axis and the normal to the lines ˚= u 2 ˇ=2 4 if d Lcos˚the needle intercepts a line (update counter N s = N s +1) 5 Repeat procedure N times 6 Estimate probability intersection P Théorème 1.2. << /S /GoTo /D (section.5.2) >> (Utilisation d'un g\351n\351rateur pseudo al\351atoire)
81 0 obj
�^���.a�t]�Խ+�-�br;��E�����������n�����w9$ ��=��h����� q�2�b�7�)��^�-��'�,À05���K��}���oo��`7�'t:�4��z�[�o�������ٵ�?��8�F'�(��?s!�?X_������]���#�~��w_�� )y9�[Ғ2�����Z�n�^$Rvq1���N_����3;;������1��t����cyzn������!�$�gFy�幤4Ъ��X���)����9����P'��[��#��a�ֶ���6�7�f�s*�#�mjl�pn)odna�`��k�����>U��M%� ��Z_z̏�X�/"1��#�?[i��V7��p��ݳ��@���Od�F4�j�~�|m���Ǝ���Y�r�W�ok��x�[�_n����������z�w�v���j���\���\�9y��S���������? endobj << /S /GoTo /D (section.6.1) >> (Contr\364le de convergence des algorithmes MCMC) (Premiers exemples) 69 0 obj 28 0 obj endobj (Exercice 2.10: Algorithme de J\366nk) endobj Monte Carlo simulation in MS Excel The Monte Carlo method is based on the generation of multiple trials to determine the expected value of a random variable. 60 0 obj Les méthode de Monte Carlo contrairement aux autres méthodes numériques reposent sur l'utilisa-tion des nombres aléatoires. << /S /GoTo /D (section.5.3) >> << /S /GoTo /D (chapter.7) >> endobj /Type /XObject endobj 3. ?y����O9t��O��?O]�")'� #%w������)�)�Z�|��������bc�u4����ͩ�9{��7,��VVW=?Xή�N��u(�W��0W½�;�)�ԒE�M,����� g�~/4e�^%�~e�9\�;8���;�Ա �Ȉ���p���SK8C[k����h����a�+��gw�^Q&|�G�IT Ʀf��F��ڦ�щ�0j��_�'�9UI6ɂ���#�#+��@6�6�s�'L�x�iQ�k�������ql�N)�. (Test d'ind\351pendance) << /S /GoTo /D (section.3.1) >> 105 0 obj (M\351thode de la transformation inverse ) 144 0 obj (Conclusion) 113 0 obj %���� endobj endobj
68 0 obj 73 0 obj /Length 82575 •Ulam is primarily known for designing the hydrogen bomb with Edward Teller in 1951. 49 0 obj
161 0 obj (Technique de conditionnement) endobj /Filter /FlateDecode 152 0 obj 104 0 obj stream (Rappels sur les chaines de Markov ) endobj
Monte Carlo simulations have been widely used by microscopists for the last few decades. 12 0 obj endobj 117 0 obj
endobj
57 0 obj
<< /S /GoTo /D (part.1) >> /SMask 198 0 R (M\351thode du rejet) 187 0 obj << 184 0 obj 176 0 obj 168 0 obj endobj endobj 97 0 obj
endobj
(Cas des variables Gaussiennes) J. STOEHR Méthodes de Monte Carlo de simulation 100 fois supérieur. 16 0 obj
(Evaluation de la variation)
endobj 160 0 obj endobj /ColorSpace /DeviceRGB
endobj endobj endobj << /S /GoTo /D (section.8.3) >> 17 0 obj
<< /S /GoTo /D (section.8.5) >> 72 0 obj Radiation Simulation and Monte Carlo Method -M. Asai (SLAC) 10 Probability Density Function (PDF) -2 • A PDF f ( x ) is a density function, i.e., it specifies the probability per unit of x , 141 0 obj (G\351n\351rateurs \340 Congruence Lin\351aire GCL) xڅQ�N�0��+����w�ʱ��Dn�!m��4"��g�(�*֞�ά=�,D赇`r��g�3�9�brÁ����Mm`�̞di����j&�t�K�����|M"m1�` �ч ^[d+�^������Ԝ��3V�ꕝ)��Ԗ=bU̶ٍ'�c#'V�����?����:H�a�n��6�n��,ij���=_�9i�^^/c��Qs��fH!��Ǣ�ʯt�;���8�V ej�ڌd�̜�^�y�ڊ?��q�`#}�ա|_� �^%҆z��|����w���آ&;e� �8�2�v�r���u�t+��1��
endobj endobj 21 0 obj (Exercice 2.10) 1.3 Different kinds of Monte Carlo simula-tions There are at least three different kinds of Monte Carlo simulations: • Transport simulations. 101 0 obj (Convergence vers la loi stationnaire.) endobj (G\351n\351rer des variables al\351atoires) (Le g\351n\351rateur de Mersenne Twister) (D\351termination de la p\351riode optimale) endobj << /S /GoTo /D (section.8.1) >> endobj %PDF-1.5 (Autres types de convergence) endobj endobj
2. endobj
<< /S /GoTo /D (subsection.8.1.2) >> endobj Monte Carlo, simulation, MCMC, estimation, optimization Abstract Many quantitative problems in science, engineering, and economics are nowadays solved via statistical sampling on a computer. endobj (Test d'uniformit\351) << /S /GoTo /D (section.4.1) >> This example differs in at least the two following ways from usual MC simulations: ... function (pdf), p(x). endobj << /S /GoTo /D (subsection.7.0.1) >>
endobj 48 0 obj a neutron) that reaches a shield.
<< /S /GoTo /D (chapter.2) >> 100 0 obj Cette méthode est intéressante puisqu'elle peut s'appliquer à des problèmes de grande dimension, comme par exemple pour calculer l'espérance de rendement sur l'ensemble d'un marché boursier. (Th\351or\350me ergodique) /Length 315 29 0 obj 65 0 obj 137 0 obj (exercices 2.07) 5 0 obj
endobj
121 0 obj 116 0 obj (Du bon choix des param\350tres) 41 0 obj (Simulation d'une loi B\352ta via Metropolis Hasting) << /S /GoTo /D (section.5.1) >> (Vitesse de convergence et temps de calcul) endobj
endobj endobj endobj (Test de Gelman Rubin) << /S /GoTo /D [186 0 R /Fit] >> endobj
186 0 obj << /Height 215 endobj
120 0 obj 33 0 obj endobj (III Simulation de Cha\356nes de Markov et algorithme de Metropolis Hastings) endobj endobj endobj << /S /GoTo /D (chapter.8) >> endobj 53 0 obj 25 0 obj endobj 85 0 obj endobj endobj 132 0 obj endobj 153 0 obj endobj
endobj << /S /GoTo /D (part.4) >>
/MediaBox [0 0 595.276 841.89] 185 0 obj endobj x��ut[W>��[��ޛ:�r�ә��N��Ift��33�efff[f�e�,[`K�l������7I�6�dS�k���V������zf�fV��7�'W�GVFGW��W�uxtevnm|j���Z���w�{��qMϬ�N������!|�68���9[�< m'?������c�}�}��s�ʣ�?8�<>�dy����������������56�6�>��m�mjjmiqknncWz2���d֤�fD�G�����؈����OOWwc����¢R_o��Ԭ���&�//o��ۧm�� b��f��s�T�疆�M�% 13 0 obj
133 0 obj (G\351n\351rer de l'al\351atoire: une n\351cessit\351 en simulation ) most MC methods are general and flexible enough to work with models of arbitrary complexity. << /S /GoTo /D (subsection.1.1.1) >>
169 0 obj
24 0 obj
endobj << /S /GoTo /D (section.8.6) >>
/Resources 190 0 R endobj endobj 108 0 obj 129 0 obj >> (M\351thode des variables antith\351tiques) 32 0 obj
stream 191 0 obj << (Cas de variables continues) (IV Exercices)
148 0 obj
(Loi conditionnelle) endobj endobj << /S /GoTo /D (section.1.2) >> 84 0 obj 172 0 obj 165 0 obj 92 0 obj endobj
Théorème central limite Soit (Xn)n‚1 une suite de variables aléatoires réelles i.i.d. endobj << /S /GoTo /D (chapter.6) >> endobj endobj For each point, determine whether it lies inside the unit circle, the red region in Figure ??.
endobj endobj endobj
endobj (Algorithme de Metropolis Hasting) << /S /GoTo /D (section.8.7) >> << /S /GoTo /D (subsection.2.1.1) >> (Test sur l'efficacit\351 des g\351n\351rateurs)
<< /S /GoTo /D (subsection.2.3.1) >>
61 0 obj 4 0 obj << /S /GoTo /D (subsection.1.2.2) >> 145 0 obj
(M\351thodes de R\351duction de la variance) (I G\351n\351rer de l'al\351atoire) << /S /GoTo /D (section.2.2) >>
128 0 obj endobj << /S /GoTo /D (section.8.4) >>
/Filter /FlateDecode 56 0 obj 45 0 obj << /S /GoTo /D (section.4.2) >>
<< /S /GoTo /D (chapter.5) >> endobj << /S /GoTo /D (chapter.4) >> << /S /GoTo /D (subsection.2.1.2) >> 109 0 obj endobj Introduction A brief overview Buffon’s experiment Monte Carlo simulation 1 Sample an u 1 ˘U[0;1) and u 2 U[0;1) 2 Calculate distance from a line: d = u 1 t 3 Calculate angle between needle’s axis and the normal to the lines ˚= u 2 ˇ=2 4 if d Lcos˚the needle intercepts a line (update counter N s = N s +1) 5 Repeat procedure N times 6 Estimate probability intersection P Théorème 1.2. << /S /GoTo /D (section.5.2) >> (Utilisation d'un g\351n\351rateur pseudo al\351atoire)
81 0 obj
�^���.a�t]�Խ+�-�br;��E�����������n�����w9$ ��=��h����� q�2�b�7�)��^�-��'�,À05���K��}���oo��`7�'t:�4��z�[�o�������ٵ�?��8�F'�(��?s!�?X_������]���#�~��w_�� )y9�[Ғ2�����Z�n�^$Rvq1���N_����3;;������1��t����cyzn������!�$�gFy�幤4Ъ��X���)����9����P'��[��#��a�ֶ���6�7�f�s*�#�mjl�pn)odna�`��k�����>U��M%� ��Z_z̏�X�/"1��#�?[i��V7��p��ݳ��@���Od�F4�j�~�|m���Ǝ���Y�r�W�ok��x�[�_n����������z�w�v���j���\���\�9y��S���������? endobj << /S /GoTo /D (section.6.1) >> (Contr\364le de convergence des algorithmes MCMC) (Premiers exemples) 69 0 obj 28 0 obj endobj (Exercice 2.10: Algorithme de J\366nk) endobj Monte Carlo simulation in MS Excel The Monte Carlo method is based on the generation of multiple trials to determine the expected value of a random variable. 60 0 obj Les méthode de Monte Carlo contrairement aux autres méthodes numériques reposent sur l'utilisa-tion des nombres aléatoires. << /S /GoTo /D (section.5.3) >> << /S /GoTo /D (chapter.7) >> endobj /Type /XObject endobj 3. ?y����O9t��O��?O]�")'� #%w������)�)�Z�|��������bc�u4����ͩ�9{��7,��VVW=?Xή�N��u(�W��0W½�;�)�ԒE�M,����� g�~/4e�^%�~e�9\�;8���;�Ա �Ȉ���p���SK8C[k����h����a�+��gw�^Q&|�G�IT Ʀf��F��ڦ�щ�0j��_�'�9UI6ɂ���#�#+��@6�6�s�'L�x�iQ�k�������ql�N)�. (Test d'ind\351pendance) << /S /GoTo /D (section.3.1) >> 105 0 obj (M\351thode de la transformation inverse ) 144 0 obj (Conclusion) 113 0 obj %���� endobj endobj
68 0 obj 73 0 obj /Length 82575 •Ulam is primarily known for designing the hydrogen bomb with Edward Teller in 1951. 49 0 obj
161 0 obj (Technique de conditionnement) endobj /Filter /FlateDecode 152 0 obj 104 0 obj stream (Rappels sur les chaines de Markov ) endobj
Monte Carlo simulations have been widely used by microscopists for the last few decades. 12 0 obj endobj 117 0 obj
endobj
57 0 obj
<< /S /GoTo /D (part.1) >> /SMask 198 0 R (M\351thode du rejet) 187 0 obj << 184 0 obj 176 0 obj 168 0 obj endobj endobj 97 0 obj
endobj
(Cas des variables Gaussiennes) J. STOEHR Méthodes de Monte Carlo de simulation 100 fois supérieur. 16 0 obj
(Evaluation de la variation)
endobj 160 0 obj endobj /ColorSpace /DeviceRGB
endobj endobj endobj << /S /GoTo /D (section.8.3) >> 17 0 obj
<< /S /GoTo /D (section.8.5) >> 72 0 obj Radiation Simulation and Monte Carlo Method -M. Asai (SLAC) 10 Probability Density Function (PDF) -2 • A PDF f ( x ) is a density function, i.e., it specifies the probability per unit of x , 141 0 obj (G\351n\351rateurs \340 Congruence Lin\351aire GCL) xڅQ�N�0��+����w�ʱ��Dn�!m��4"��g�(�*֞�ά=�,D赇`r��g�3�9�brÁ����Mm`�̞di����j&�t�K�����|M"m1�` �ч ^[d+�^������Ԝ��3V�ꕝ)��Ԗ=bU̶ٍ'�c#'V�����?����:H�a�n��6�n��,ij���=_�9i�^^/c��Qs��fH!��Ǣ�ʯt�;���8�V ej�ڌd�̜�^�y�ڊ?��q�`#}�ա|_� �^%҆z��|����w���آ&;e� �8�2�v�r���u�t+��1��