Upgrade to Pro — share decks privately, control downloads, hide ads and more …

感染症の数理モデル3

 感染症の数理モデル3

Daisuke Yoneoka

March 22, 2024
Tweet

More Decks by Daisuke Yoneoka

Other Decks in Research

Transcript

  1. ⽬次 1. 感染症のコンパートメントモデル 2. 基本再⽣産数 3. 最終流⾏規模 4. R実装 5.

    ⼈⼝の異質性とSIR 6. 再⽣産⽅程式とエボラ vs インフル 本書の内容をカバーします。 具体的なコードなどは右の本 詳細なプログラムなどは https://github.com/objornstad/epimdr/tree/ master/rcode (結構間違ってる。。。) 2/48
  2. エンデミックSIR 単位時間あたりb⼈誕⽣ 感受性者も感染者も回復者も、単位時間当たり⼀定の割合μ(⼈/単位時間) で (その感染症でない理由によって) 死亡(感染症が理由でないことを表すため「⾃然死」とも呼ぶ) 16 <latexit sha1_base64="NvWFqkTvyTGUqoy+uW80gQOtDHA=">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</latexit> dS(t)

    dt = b µS(t) S(t)I(t) dI(t) dt = S(t)I(t) ( + µ)I(t) dR(t) dt = I(t) µR(t) <latexit sha1_base64="UuVs7790AhEJPpVkp7fKbSMNZCU=">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</latexit> dN(t) dt = b µ(S(t) + I(t) + R(t)) = 0 ) N(t) = b µ N(t) = S(t)+I(t)+R(t)はb/μを安定な平衡値としてもつ ので、S, I, Rの2つのうち⼀つが決まれば後は決まる ので、上2つで⼗分じゃん ⾏列表記しておこう <latexit sha1_base64="ziF6xaENJaT0Tp2RdtM46DBREzM=">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</latexit> x(t) = ✓ S(t) I(t) ◆ <latexit sha1_base64="HGe8DoBfD/ZpFEIRM5/3fmVSQ8o=">AAACqXicfVBNbxMxEHWWr7J8pXDkYhEhpbSNdhEtHECqgAM5IFqFtBFxFM06s1ur/ljZXtRolV/GL+HIFf4E3jRFtEWMZOnpvRnPvJeVUjifJN9b0bXrN27eWrsd37l77/6D9vrDQ2cqy3HIjTR2lIFDKTQOvfASR6VFUJnEo+zkXaMffUXrhNGf/bzEiYJCi1xw8IGatod5d7BF+xv0DWUZFkLXpQJvxekizraZquiAbjeKBzroMxafw8B2g7zJClAKNvoxQz37Mzttd5Jesix6FaQr0CGr2p+ut16zmeGVQu25BOfGaVL6SQ3WCy5xEbPKYQn8BAocB6hBoZvUS/8L+jQwM5obG572dMn+PVGDcm6ustAZDjx2l7WG/Jc2rnz+alILXVYeNT9blFeSekObMOlMWORezgMAbkW4lfJjsMB9iPzCluZvb4x0wcp7DBYtfgzUpxIteGOf1QxsoYReBMsF22rQ/xrh9LwxoDgOeaeX070KDp/30t3ezsGLzt7bVfJr5DF5QrokJS/JHvlA9smQcPKN/CA/ya9oMzqIRtGXs9aotZp5RC5UxH8DNX/Rfw==</latexit> f(S, I) = ✓ b µS SI SI (µ + )I ◆ とおくと、上2つの式は <latexit sha1_base64="/BZRLvoWQ0SKf5c3bLlt0XMTM9M=">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</latexit> dx(t) dt = f(x(t)) よく使うので、ヤコビアンも定義しとく <latexit sha1_base64="uxtQikUhlTJa0XzJc+mR1VSqhC0=">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</latexit> f0(S, I) = ✓ µ I S I S (µ + ) ◆
  3. ワクチンの効果 eの割合だけワクチン接種している = R(0)>0 (免疫をつけている) cf. (さっきまでは(S(0), I(0),R(0))=(N,0,0)だったけど、今は (S(0), I(0),R(0))=((1-e)N,0,eN))

    22 <latexit sha1_base64="oMr7G/svA9fDN76DQkTXeU9g4jE=">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</latexit> dS(t) dt = b(1 e) µS(t) S(t)I(t) dI(t) dt = (µ + )I(t) + S(t)I(t) dR(t) dt = be µR(t) I(t) ((1-e)N,0,eN)) = (b(1-e)/μ, 0, be/μ) を安定な平衡値としてもつ このとき少数の感染者が⼊ってきたら、再⽣産過程は以下 <latexit sha1_base64="InRFdEtl5ZpoewqBvfuaDU8L5JY=">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</latexit> dI(t) dt = (µ + )I(t) + b(1 e) µ I(t) > 0 <latexit sha1_base64="F7m6a5Ev4Axl7ctn8T9eiPzISqo=">AAACIHicbVDLSgMxFM3UV62vqks3wSK0qGVGiroRim5c1mIf0Cklk95pQ5OZIckIZeiPuPRLXKobceFCv8b0sdDWAxcO55xLco8Xcaa0bX9ZqaXlldW19HpmY3Nreye7u1dXYSwp1GjIQ9n0iALOAqhppjk0IwlEeBwa3uBm7DceQCoWBvd6GEFbkF7AfEaJNlInW6rmoYCvsNv1JaGJ64Em2MN55xQKo8QVcd7MseEmMxGrHRt3sjm7aE+AF4kzIzk0Q6WT/XS7IY0FBJpyolTLsSPdTojUjHIYZdxYQUTogPSgZWhABKh2MrluhI+M0sV+KM0EGk/U3xsJEUoNhXfiCRMWRPfVvD0W//NasfYv2wkLolhDQKdv+THHOsTjtnCXSaCaDw0hVDLzXUz7xDSlTacZ04Mzf/UiqZ8VnfNi6a6UK1/PGkmjA3SI8shBF6iMblEF1RBFj+gZvaI368l6sd6tj2k0Zc129tEfWN8/I+ifNw==</latexit> R(e) = b(1 e) µ(µ + e) = (1 e)R0 実効再⽣産数 <latexit sha1_base64="uMrhJGBeAXrqFujfAKJogCUqZoM=">AAACAHicbVDLSsNAFJ3UV62v+Ni5GSyCCy2JFHUjFN24rGIf0IYwmU7aoTOTMDMRasjGL3GpbsSt/+HCv3HaZqGtBy4czrmXe+8JYkaVdpxvq7CwuLS8Ulwtra1vbG7Z2ztNFSUSkwaOWCTbAVKEUUEammpG2rEkiAeMtILh9dhvPRCpaCTu9SgmHkd9QUOKkTaSb+8ReAndk24vlAinbpbe+U7m22Wn4kwA54mbkzLIUfftr24vwgknQmOGlOq4Tqy9FElNMSNZqZsoEiM8RH3SMVQgTpSXTq7P4KFRejCMpCmh4UT9PZEirtSIB8cBN80c6YGatcfif14n0eGFl1IRJ5oIPN0VJgzqCI7TgD0qCdZsZAjCkppzIR4gE4U2mZVMDu7s1/OkeVpxzyrV22q5dpUnUgT74AAcARecgxq4AXXQABg8gmfwCt6sJ+vFerc+pq0FK5/ZBX9gff4Ao3eVJw==</latexit> e = 1 1 R0 R(e)<1であればよいので、 臨界免疫化割合 これより⼤きいと集団免疫
  4. 再⽣産⽅程式 (renewal equation) 再⽣産⽅程式:感染ダイナミクスを簡略化した SIR モデルではなく、 もう⼀つの重要なアイディア 再⽣過程を記述する⽅程式の⼀種 時刻 tにおける新規感染者数

    I(t) は、a⽇前の過去の新規感染者数を I(t−a)を⽤いて 基本再⽣産数は、全感染性期間を積分することで 23/20 <latexit sha1_base64="KRYtOBHRA80qCdkLhBLyz4qB/bc=">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</latexit> I(t) = Z 1 0 A(a)I(t a)da A(a)は、⼀⼈の感染者がある時刻で感染した場合、そのa⽇後に平均的に感染させる⼈数を表現。 (A(a)が正規化されている場合)A(a)がserial interval (⼆次感染までの時間)の確率密度関数となる!!! <latexit sha1_base64="vFZ72HbnVeU8w6LkYvqK9aOFGaY=">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</latexit> R0 = Z 1 0 A(a)da
  5. 実際の解き⽅ 特性曲線法を使って頑張って解く(12、13室は演習問題?) 25/20 <latexit sha1_base64="icnoX6AznnzM97ilqbsTvRLdUjU=">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</latexit> I(t, 0) = S(0) 8

    < : Z t 0 (a) exp ✓ Z a 0 ( )d ◆ I(t a, 0)da + Z 1 t (a) exp R a 0 ( )d exp ⇣ R a t 0 ( )d ⌘I(0, a t)da 9 = ; I(0, a-t)は初期t=0の感染者がどれだけ 感染させたかを表現。つまり、tがある 程度⻑いと0であることが期待される <latexit sha1_base64="GTRq6qknRaiXtLrwk85pERQHa50=">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</latexit> I(t, 0) = Z 1 0 A(a)I(t a, 0)da 第⼀項⽬だけ <latexit sha1_base64="smcD8zjsjLXXsER4pdJiE6kbOOA=">AAACmnicfVDbbhMxEHWWWwm3FB7hwSJCSlCJdhG3F0RpeQAhRBFNWykO0awzu7Fqr1f2LGq0ygfxNbzSv8GbBIm2iJEsH585Y/uctNTKUxyftqJLl69cvbZxvX3j5q3bdzqbdw+8rZzEobTauqMUPGpV4JAUaTwqHYJJNR6mx7tN//A7Oq9ssU/zEscG8kJlSgIFatLZfduDPn/Nv/bivkiRIBwFnpRCY0a9J0IVNIm/ARc5GAM94VVuoD9d7cKpfEb9SacbD+Jl8YsgWYMuW9feZLP1RkytrAwWJDV4P0riksY1OFJS46ItKo8lyGPIcRRgAQb9uF66XfBHgZnyzLqwCuJL9u+JGoz3c5MGpQGa+fO9htxKzb/ao4qyV+NaFWVFWMjVW1mlOVnepMenyqEkPQ8ApFPhu1zOwIGkkPGZh5q7yVrtg5t3GFw6/BSozyU6IOse1wJcblSxCK5zsdWg/wnh5I8woHY7RJ6cD/giOHg6SF4Mnn951t3eWYe/we6zh6zHEvaSbbP3bI8NmWQ/2E/2i51GD6Kd6EP0cSWNWuuZe+xMRfu/AXYtzK0=</latexit> A(a) = S(0) (a) exp ✓ Z a 0 ( )d ◆ ただし <latexit sha1_base64="QvFYNraTbOF3WY09zCRxI8hPGLg=">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</latexit> A(a) R 1 0 A(a)da 再⽣産核 は⼆次感染までの時間(serial interval) の確率密度関数とみなせる
  6. エボラは感染しやすいのか?(1) エボラウイルス病のR0 =1.5 ~ 2.0 H1N1-2009のR0 = 1.4 26/20 「エボラとインフルの感染性は同じくらいだ」

    というstatementは正しいか? <latexit sha1_base64="KRYtOBHRA80qCdkLhBLyz4qB/bc=">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</latexit> I(t) = Z 1 0 A(a)I(t a)da の解を と⼀旦してみる <latexit sha1_base64="g1gRI6Y400KhcDNHvGyOUuELeYQ=">AAAB/XicbVBNS8NAEN34WetXqkcvi0VoQUoiRb0IRS96q2A/oA1ls920SzebsDtRSyn+Eo/qRbz6Szz4b9y2OWjrg4HHezPMzPNjwTU4zre1tLyyurae2chubm3v7Nq5vbqOEkVZjUYiUk2faCa4ZDXgIFgzVoyEvmANf3A18Rv3TGkeyTsYxswLSU/ygFMCRurYuZsCFPEFHrTZY1xQGIodO++UnCnwInFTkkcpqh37q92NaBIyCVQQrVuuE4M3Igo4FWycbSeaxYQOSI+1DJUkZNobTU8f4yOjdHEQKVMS8FT9PTEiodbD0D/2Q9McEujreXsi/ue1EgjOvRGXcQJM0tmuIBEYIjyJAne5YhTE0BBCFTfnYtonilAwgWVNDu7814ukflJyT0vl23K+cpkmkkEH6BAVkIvOUAVdoyqqIYoe0DN6RW/Wk/VivVsfs9YlK53ZR39gff4AmWaTYQ==</latexit> I(t) = k exp(rt) <latexit sha1_base64="YqBVCmEG4Xltxg7eyG3luZYdW48=">AAACG3icbVBNSwMxEM36bf2qevQSLEILWnelqBfBj4vHCrYKbS2zabYNzWaXZFYsxZ/h0V/iUb2IVw/+G9N2D9r6YJjHezMk8/xYCoOu++1MTc/Mzs0vLGaWlldW17LrG1UTJZrxCotkpG99MFwKxSsoUPLbWHMIfclv/O7FwL+559qISF1jL+aNENpKBIIBWqmZ3e/W+UOc1xQL9ITWhcKmS+9sD7BHz/JQSP087kGh0IJmNucW3SHoJPFSkiMpys3sV70VsSTkCpkEY2qeG2OjDxoFk/wxU08Mj4F1oc1rlioIuWn0h4c90h2rtGgQaVsK6VD9vdGH0Jhe6O/6oR0OATtm3B6I/3m1BIPjRl+oOEGu2OitIJEUIzoIiraE5gxlzxJgWtjvUtYBDQxtnBmbgzd+9SSpHhS9w2LpqpQ7PU8TWSBbZJvkiUeOyCm5JGVSIYw8kRfyRt6dZ+fV+XA+R6NTTrqzSf7A+foBwBeeuw==</latexit> k exp(rt) = Z 1 0 A(a)k exp(r(t a))da でないといけない。両辺 で割れば <latexit sha1_base64="g1gRI6Y400KhcDNHvGyOUuELeYQ=">AAAB/XicbVBNS8NAEN34WetXqkcvi0VoQUoiRb0IRS96q2A/oA1ls920SzebsDtRSyn+Eo/qRbz6Szz4b9y2OWjrg4HHezPMzPNjwTU4zre1tLyyurae2chubm3v7Nq5vbqOEkVZjUYiUk2faCa4ZDXgIFgzVoyEvmANf3A18Rv3TGkeyTsYxswLSU/ygFMCRurYuZsCFPEFHrTZY1xQGIodO++UnCnwInFTkkcpqh37q92NaBIyCVQQrVuuE4M3Igo4FWycbSeaxYQOSI+1DJUkZNobTU8f4yOjdHEQKVMS8FT9PTEiodbD0D/2Q9McEujreXsi/ue1EgjOvRGXcQJM0tmuIBEYIjyJAne5YhTE0BBCFTfnYtonilAwgWVNDu7814ukflJyT0vl23K+cpkmkkEH6BAVkIvOUAVdoyqqIYoe0DN6RW/Wk/VivVsfs9YlK53ZR39gff4AmWaTYQ==</latexit> I(t) = k exp(rt) オイラー- トロカの特性⽅程式 <latexit sha1_base64="J/e5A2LBdnws9TuyTGuk9tcHYis=">AAACDnicbVDLSgNBEJz1GeMr6tHLkCAkoGFXRL0IUS8eI5gHJDH0Tmbj4OzsMtMrhpC7R7/Eo3oRr/6AB//GyeOg0YKmi6puZrr8WAqDrvvlzMzOzS8sppbSyyura+uZjc2qiRLNeIVFMtJ1HwyXQvEKCpS8HmsOoS95zb89H/q1O66NiNQV9mLeCqGrRCAYoJXamaxHT2hTKGy79Nr2AHv0NA+FJr+P83saCh1oZ3Ju0R2B/iXehOTIBOV25rPZiVgScoVMgjENz42x1QeNgkk+SDcTw2Ngt9DlDUsVhNy0+qNbBnTHKh0aRNqWQjpSf270ITSmF/q7fmiHQ8AbM20Pxf+8RoLBcasvVJwgV2z8VpBIihEdZkM7QnOGsmcJMC3sdym7AQ0MbYJpm4M3ffVfUt0veofFg8uDXOlskkiKbJMsyROPHJESuSBlUiGMPJAn8kJenUfn2Xlz3sejM85kZ4v8gvPxDWZHmc8=</latexit> 1 = Z 1 0 A(a) exp( ra)da <latexit sha1_base64="QvFYNraTbOF3WY09zCRxI8hPGLg=">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</latexit> A(a) R 1 0 A(a)da ここで がserial intervalのpdfであることを思い出すと <latexit sha1_base64="DDEx2FgrbA2KW0FzqNF85JdxCjA=">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</latexit> R0 = 1 Z 1 0 A(a) R 1 0 A(b)db exp( ra)da この傾きが増加率r→
  7. エボラは感染しやすいのか?(2) ⿇疹 R0 = 15.0 エボラ R0 =1.4 インフル R0

    = 1.7 Serial intervalの平均値 ⿇疹 = 12⽇ エボラ = 15⽇ インフル = 3⽇ 27/20 <latexit sha1_base64="DDEx2FgrbA2KW0FzqNF85JdxCjA=">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</latexit> R0 = 1 Z 1 0 A(a) R 1 0 A(b)db exp( ra)da ↑R0 はserial intervalと増殖率rに 依存していることがわかる 右式より増殖率r が計算可能 (図2.4) ポイント:R0は「世代」ごとの感染性の指標 エボラは世代交代に2週間 vs インフルは3⽇ →インフルの⽅が感染性が⾼い →逆にエボラは隔離などで時間を稼ぎやすい
  8. Nishiura et al. (2016, EID)(1) 潜伏期間の2倍とかって基準は不確実性を⼊れてないよね Branching processに基づいて近似計算(contact tracingのデータなどは使っていないとい う意味において)

    【準備】 • Serial intervalの累積密度分布 (ガンマ分布) • ⼀⼈の⼈が何⼈に移すかの分布 (negative binomial分布) • R0 = 0.75, k = 0.14 30/20 <latexit sha1_base64="47yYbLZmmQtK0XddAAHV9JDnsas=">AAACWHicbVFdT9swFHXCRqF8lfK4F2/VpFaFKkEV8IKE2AN77Ka1gJpSOa7TWrGTyL5BRFH+JBIP7JfsEbfNNEa5kq3jc87VtY/9RHANjvNs2WsfPq5XNjarW9s7u3u1/fpAx6mirE9jEasbn2gmeMT6wEGwm0QxIn3Brv3w21y/vmdK8zj6BVnCRpJMIx5wSsBQ45roNW/PsxY+x94kUITm3hWRkjTD9kOryLPPf4+twhMsgGbp+jl2ivnWDgtP8ekMWnfZ0uC2X1v+yflRWIxrDafjLAqvArcEDVRWb1x78iYxTSWLgAqi9dB1EhjlRAGnghVVL9UsITQkUzY0MCKS6VG+SKXAXw0zwUGszIoAL9jXHTmRWmfSP/SlMUsCM/1WnpPvacMUgrNRzqMkBRbR5awgFRhiPE8ZT7hiFERmAKGKm+tiOiMmFjB/UTU5uG9fvQoGxx33pNP90W1cXJaJbKBP6AtqIhedogv0HfVQH1H0iP5YlmVbv21kV+zNpdW2yp4D9F/Z9RdcGrKs</latexit> P(Y = y) = (k + x) y! (k) ✓ R0 R0 + k ◆y ✓ 1 + R0 k ◆ k 平均: R0 分散: R0 +R0 2/k この形がover-dispersionを表現している この分布の形がf(x) <latexit sha1_base64="OOmQKJP9zanNk9QB69eitBTKIJE=">AAACDHicbZDLSgMxFIYz9VbrbdSlm2ARWpAyI0VdKBQFcVmhN2hryaSZNjSTGZIz0lK6demTuFQ34tY3cOHbmF4W2vpD4OM/53Byfi8SXIPjfFuJpeWV1bXkempjc2t7x97dq+gwVpSVaShCVfOIZoJLVgYOgtUixUjgCVb1etfjevWBKc1DWYJBxJoB6Ujuc0rAWC0b32Qgiy9xMVO6mECDS2g594D9TD+L2/2WnXZyzkR4EdwZpNFMxZb91WiHNA6YBCqI1nXXiaA5JAo4FWyUasSaRYT2SIfVDUoSMN0cTi4Z4SPjtLEfKvMk4In7e2JIAq0HgXfsBaY5INDV8+Wx+V+tHoN/3hxyGcXAJJ3u8mOBIcTjZHCbK0ZBDAwQqrj5LqZdoggFk1/K5ODOX70IlZOce5rL3+XThatZIkl0gA5RBrnoDBXQLSqiMqLoET2jV/RmPVkv1rv1MW1NWLOZffRH1ucPShyX/g==</latexit> F(t) = P(T < t) = Z t 0 f(x)dx
  9. Nishiura et al. (2016, EID)(2) 今⼿元には、N⼈の発症⽇データti がある状況 ある⽇tにoutbreakが終わっている確率は以下で(近似的に)表現できる もし1⼈だけにうつしたなら: もし2⼈にうつしたなら:

    もし2⼈にt1 とt2 の時間にうつしたなら: 31/20 <latexit sha1_base64="RHfTVsPOuTH7o8WfyEtzHS91f1U=">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</latexit> P(outbreak over at day t) = N Y i=1 1 X y=0 P(Y = y)F(t ti)y <latexit sha1_base64="+2O5TaslSr2FsDonIiuJG7a+xc4=">AAACFHicbVDLSgMxFM3UV62vUZdugkVoQcqMFHUjFAVxWcE+oB1KJk3b0MxkSO6Iw9BfcOmXuFQ34tadC//G9LHQ1gOBwznncnOPHwmuwXG+rczS8srqWnY9t7G5tb1j7+7VtYwVZTUqhVRNn2gmeMhqwEGwZqQYCXzBGv7wauw37pnSXIZ3kETMC0g/5D1OCRipYxeqhTawB0hlDL4ZHGJp4pgA7pIEjzAU8QW+LkCxY+edkjMBXiTujOTRDNWO/dXuShoHLAQqiNYt14nAS4kCTgUb5dqxZhGhQ9JnLUNDEjDtpZOLRvjIKF3ck8q8EPBE/T2RkkDrJPCP/cCEAwIDPW+Pxf+8Vgy9cy/lYRQDC+l0Vy8WGCQeN4S7XDEKIjGEUMXNdzEdEEUomB5zpgd3/upFUj8puael8m05X7mcNZJFB+gQFZCLzlAF3aAqqiGKHtEzekVv1pP1Yr1bH9NoxprN7KM/sD5/APvwnNM=</latexit> P(outbreak over at day t) = F(t) <latexit sha1_base64="c2V16gkeeDG6CvfbFi9yE96TPLQ=">AAACGHicbVDLSgMxFM34rPVVdekmWIQWSpkpRd0IRUFcVrAPaMeSSdM2NDMZkjviMPQnXPolLtWNuBVc+Demj4W2HggczjmXm3u8UHANtv1tLS2vrK6tpzbSm1vbO7uZvf26lpGirEalkKrpEc0ED1gNOAjWDBUjvidYwxtejv3GPVOay+AW4pC5PukHvMcpASN1MoVqrg3sARIZgWcGh1iaOCaAuyTGIwx5fI5zVznI5+9KnUzWLtoT4EXizEgWzVDtZL7aXUkjnwVABdG65dghuAlRwKlgo3Q70iwkdEj6rGVoQHym3WRy1QgfG6WLe1KZFwCeqL8nEuJrHftewfNN2Ccw0PP2WPzPa0XQO3MTHoQRsIBOd/UigUHicUu4yxWjIGJDCFXcfBfTAVGEgukybXpw5q9eJPVS0Tkplm/K2crFrJEUOkRHKIccdIoq6BpVUQ1R9Iie0St6s56sF+vd+phGl6zZzAH6A+vzBx/lndw=</latexit> P(outbreak over at day t) = (F(t))2 <latexit sha1_base64="y0wLlvwzQqxLuLw6YXFtU/cQzaw=">AAACHHicbVDLSgMxFM34rPU16tJNsAgtSJkpRd0IRUFcVrAPaIeSSdM2NDMZkjviMPQ3XPolLtWNuFTwb0wfC229kHA4D5J7/EhwDY7zbS0tr6yurWc2sptb2zu79t5+XctYUVajUkjV9IlmgoesBhwEa0aKkcAXrOEPr8Z6454pzWV4B0nEvID0Q97jlIChOrZTzbeBPUAqY/BNcIilsWMCuEsSPMJQwBf4Og8dtzC+S4WOnXOKzmTwInBnIIdmU+3Yn+2upHHAQqCCaN1ynQi8lCjgVLBRth1rFhE6JH3WMjAkAdNeOtlshI8N08U9qcwJAU/Y34mUBFongX/iB8YcEBjoeXlM/qe1YuideykPoxhYSKdv9WKBQeJxU7jLFaMgEgMIVdx8F9MBUYSC6TNrenDnt14E9VLRPS2Wb8u5yuWskQw6REcoj1x0hiroBlVRDVH0iJ7RK3qznqwX6936mFqXrFnmAP0Z6+sH7J6fTw==</latexit> P(outbreak over at day t) = F(t1)F(t2) ⼀般化 これを尤度とみて、y ~ NBからサンプルingすればベイズ的な推論が可能 でも、yがNBでFがガンマとかなら解析的にかけると思う
  10. Nishiura et al. (2016)の拡張について 32 Bradbury et al. (2020, Interface)

    Contact tracingで誰→誰っていう情報全く使 ってないので、使おう Linton et al. (2021,IJID) 報告遅れの要素⼊れておこう F()に畳み込みをする ぶっちゃけ、まだまだ⾊々拡張できそう。