清水_backup の履歴(No.45)

	月曜日	火曜日	水曜日	木曜日	金曜日
1-2
3-4		電波工学特論		人間情報工学
昼休み
5-6					科学技術論
7-8
9-10
11-12

個人情報系

https://www.kikn.fms.meiji.ac.jp/paper/2018/master/ito/hakuron_ITO.pdf

https://www.mext.go.jp/component/a_menu/education/micro_detail/__icsFiles/afieldfile/2018/08/22/1408402_1.pdf

暗号化データベース上のk-匿名化技術の提案

https://www.jstage.jst.go.jp/article/jjsai/24/2/24_283/_pdf

https://www.soumu.go.jp/main_content/000462276.pdf

https://shinmeisha.co.jp/newsroom/2020/02/03/%E3%82%B2%E3%83%BC%E3%83%88%E3%82%A6%E3%82%A7%E3%82%A4%E3%81%A8%E3%81%AF/

https://www.jstage.jst.go.jp/article/mjir/7/0/7_100/_pdf/-char/ja

https://qiita.com/gg_hatano/items/cf87fe5e1166a2ba80b2

https://dspace.jaist.ac.jp/dspace/bitstream/10119/17089/5/paper.pdf

https://www.jstage.jst.go.jp/article/isciesci/63/2/63_51/_pdf/-char/ja

https://qiita.com/saltcooky/items/3b5537170576c917a4b9#%E5%B7%AE%E5%88%86%E3%83%97%E3%83%A9%E3%82%A4%E3%83%90%E3%82%B7%E3%83%BC

差分プライバシー

https://www.jstage.jst.go.jp/article/jjet/41/3/41_42028/_pdf

https://www.stat.go.jp/training/2kenkyu/research/pdf/resear39.pdf

https://www.nstac.go.jp/sys/files/static/services/society_paper/30_08_02.pdf

https://www.slideshare.net/NobuhikoKondo/ir-79402108

https://www.ism.ac.jp/editsec/toukei/pdf/51-2-321.pdf

インピーダンス算出（最後のコメントアウトで各力を切り替え）
import numpy as np
from numpy.random import *
from scipy.stats import multivariate_normal
import matplotlib.pyplot as plt
import codecs
import os
import pandas as pd
import csv
import scipy
import warnings
import sympy
import matplotlib.patches as patches
from scipy.linalg import eig
from scipy.sparse.linalg import eigs
from sklearn import linear_model
from pandas import Series,DataFrame
from decimal import Decimal

np.set_printoptions(suppress = True,precision=4)

for j in range(10):
    wine = pd.read_csv("kawa"+str(j+1)+".csv", sep=",", header = None)
    a = len(wine)
    wine[6]  = 0.0

aa = [] 
    bb = []
    cc = []
    dd = []
    xx = []
    yy = []
    """
    if j == 0:
        a1 = -10802.939439
        a2 = -354.603396
        a3 = -416.326051
        a4 = -10240.90197
        b1 = -124.615634
        b2 = 0.376629
        b3 = -0.196127
        b4 = -125.146711
        s1 = -197.60076131137947
        s2 = -2.6388241164378803
        
    elif j == 1:
        a1 = -12185.956220
        a2 = -276.177196
        a3 = 64.891629
        a4 = -11143.257732
        b1 = -124.578583
        b2 = -0.070172
        b3 = 0.105672
        b4 = -124.636440
        s1 = -104.31846365812942
        s2 = 56.29908001650746
        
    elif j == 2:
        a1 = -10697.653716
        a2 = -425.032375
        a3 = -448.590504
        a4 = -10923.153243
        b1 = -124.578583
        b2 = 0.043238
        b3 = 0.20388
        b4 = -124.959393
        s1 = 111.48863433775732
        s2 = 107.75472578596722
        
    elif j == 3:
        a1 = -10790.741955
        a2 = -325.288641
        a3 = -727.872816
        a4 = -11519.279826
        b1 = -124.852590
        b2 = -0.095973
        b3 = 0.092372
        b4 = -125.003093
        s1 = -136.28863741882049
        s2 = -41.591209849403235
        
    elif j == 4:
        a1 = -11346.258562
        a2 = -1.149710
        a3 = -93.341148
        a4 = -13413.429431
        b1 = -124.702401
        b2 = 0.031475
        b3 = -0.436953
        b4 = -124.978885
        s1 = 134.3578611727106
        s2 = -24.86550420729226
        
    elif j == 5:
        a1 = -10854.110140
        a2 = -4.921194
        a3 = -175.519349
        a4 = -10095.219348
        b1 = -125.222402
        b2 = 0.227140
        b3 = -0.769586
        b4 = -124.439348
        s1 = 145.7122653453931
        s2 = 300.4932749831406
        
    elif j == 6:
        a1 = -10209.866654
        a2 = -202.290530
        a3 = -338.165905
        a4 = -12105.504932
        b1 = -124.622130
        b2 = -0.157876
        b3 = 0.105802
        b4 = -124.756109
        s1 = 188.0537855782985
        s2 = 148.61675577909799
        
    elif j == 7:
        a1 = -10401.300906
        a2 = -338.489043
        a3 = -156.009696
        a4 = -10144.294145
        b1 = -124.961672
        b2 = 0.133852
        b3 = -0.113044
        b4 = -124.701182
        s1 = -189.10578732712446
        s2 = -41.54183223927709
        
    elif j == 8:
        a1 = -9763.092446
        a2 = 107.655850
        a3 = 137.294048
        a4 = -9216.590263
        b1 = -125.246290
        b2 = 0.441718
        b3 = -0.638424
        b4 = -124.251535
        s1 = -368.8550697975761
        s2 = -518.6014077997578

elif j == 9: 
        a1 = -9137.638545
        a2 = 168.192283
        a3 = 95.770789
        a4 = -87900.990643
        b1 = -124.584943
        b2 = -0.226311
        b3 = -0.116026
        b4 = -124.755293
        s1 = 302.4395296188067
        s2 = -134.2139800672458
        """
    if j == 0:
        a1 = -10802.939439
        a2 = 0
        a3 = 0
        a4 = -10240.90197
        b1 = -124.615634
        b2 = 0
        b3 = 0
        b4 = -125.146711
        s1 = -197.60076131137947
        s2 = -2.6388241164378803
        
    elif j == 1:
        a1 = -12185.956220
        a2 = 0
        a3 = 0
        a4 = -11143.257732
        b1 = -124.578583
        b2 = 0
        b3 = 0
        b4 = -124.636440
        s1 = -104.31846365812942
        s2 = 56.29908001650746
        
    elif j == 2:
        a1 = -10697.653716
        a2 = 0
        a3 = 0
        a4 = -10923.153243
        b1 = -124.578583
        b2 = 0
        b3 = 0
        b4 = -124.959393
        s1 = 111.48863433775732
        s2 = 107.75472578596722
        
    elif j == 3:
        a1 = -10790.741955
        a2 = 0
        a3 = 0
        a4 = -11519.279826
        b1 = -124.852590
        b2 = 0
        b3 = 0
        b4 = -125.003093
        s1 = -136.28863741882049
        s2 = -41.591209849403235
        
    elif j == 4:
        a1 = -11346.258562
        a2 = 0
        a3 = 0
        a4 = -13413.429431
        b1 = -124.702401
        b2 = 0
        b3 = 0
        b4 = -124.978885
        s1 = 134.3578611727106
        s2 = -24.86550420729226
        
    elif j == 5:
        a1 = -10854.110140
        a2 = 0
        a3 = 0
        a4 = -10095.219348
        b1 = -125.222402
        b2 = 0
        b3 = 0
        b4 = -124.439348
        s1 = 145.7122653453931
        s2 = 300.4932749831406
        
    elif j == 6:
        a1 = -10209.866654
        a2 = 0
        a3 = 0
        a4 = -12105.504932
        b1 = -124.622130
        b2 = 0
        b3 = 0
        b4 = -124.756109
        s1 = 188.0537855782985
        s2 = 148.61675577909799
        
    elif j == 7:
        a1 = -10401.300906
        a2 = 0
        a3 = 0
        a4 = -10144.294145
        b1 = -124.961672
        b2 = 0
        b3 = 0
        b4 = -124.701182
        s1 = -189.10578732712446
        s2 = -41.54183223927709
        
    elif j == 8:
        a1 = -9763.092446
        a2 = 0
        a3 = 0
        a4 = -9216.590263
        b1 = -125.246290
        b2 = 0
        b3 = 0
        b4 = -124.251535
        s1 = -368.8550697975761
        s2 = -518.6014077997578

elif j == 9: 
        a1 = -9137.638545
        a2 = 0
        a3 = 0
        a4 = -87900.990643
        b1 = -124.584943
        b2 = 0
        b3 = 0
        b4 = -124.755293
        s1 = 302.4395296188067
        s2 = -134.2139800672458
    for i in range(a):
        if i == 0:
            aa.append(float(wine[0][i+1] - wine[0][i]))
            bb.append(float(wine[1][i+1] - wine[1][i]))
            #aa.append(float(wine[4][i] - float(wine[0][i])))
            #bb.append(float(wine[4][i] - float(wine[1][i])))
            cc.append(float(wine[4][i+1]))
            dd.append(float(wine[5][i+1]))

elif i == a -1:
            aa.append(float(wine[0][i] - wine[0][i]))
            bb.append(float(wine[1][i] - wine[1][i]))
            #aa.append(float(wine[4][i] - float(wine[0][i])))
            #bb.append(float(wine[4][i] - float(wine[1][i])))
            cc.append(float(0))
            dd.append(float(0))

else:
            aa.append(float(wine[0][i+1] - wine[0][i]))
            bb.append(float(wine[1][i+1] - wine[1][i]))
            #aa.append(float(wine[4][i] - float(wine[0][i])))
            #bb.append(float(wine[4][i] - float(wine[1][i])))
            cc.append(float(wine[4][i+1]))
            dd.append(float(wine[5][i+1]))
    wine[7] = aa
    wine[8] = bb
    xx = -wine[2]
    yy = -wine[3]
    wine[9] =xx
    wine[10] =yy
    #wine[11] = 1.0
    #wine[12] = 1.0
    wine[11] = cc
    wine[12] = dd
    #wine[13] = wine[4]

df = wine.loc[:,7:12]
 
    
    B = 0
    K = 0
    gx = 0
    gy = 0
    
    for k in range(a):
        ex = a1*aa[k] + a2*bb[k] + b1*xx[k] + b2*yy[k] + s1
        ey = a3*aa[k] + a4*bb[k] + b3*xx[k] + b4*yy[k] + s2
        
        #c1
        gx =s1 + (cc[k] - ex) 
        #c2
        gy =s2 + (dd[k] - ey

x/a,0.],\n [0.,gy/a]\n ])) \n\n #print(M)\n K = np.dot(M , np.array([\n [a1,a2],\n [a3,a4]\n ]))\n\n B = np.dot(M , np.array([\n [b1,b2],\n [b3,b4]\n ]))\n\n\n #print(str(j+1)+"回目の慣性行列")\n #print(M)\n #print(str(j+1)+"回目の粘性行列")\n #print(B)\n print(str(j+1)+"回目の剛性行列")\n print(K))

内部モデルの信頼度の算出
import numpy as np
from numpy.random import *
from scipy.stats import multivariate_normal
import matplotlib.pyplot as plt
import codecs
import os
import pandas as pd
import csv
import scipy
import warnings
import sympy
import matplotlib.patches as patches
import math
from scipy.linalg import eig
from scipy.sparse.linalg import eigs
from sklearn import linear_model
from pandas import Series,DataFrame

np.set_printoptions(suppress = True)
warnings.filterwarnings('ignore')

# Axesを配置する領域を作成
#fig = plt.figure(facecolor = "lightgray")
#fig = plt.figure(figsize=(5,5))

# FigureにAxesを追加
#ax = fig.add_subplot(111)

#r = patches.Rectangle(xy=(120, 120), width=-240., height=-240., ec='#000000', fill=False)
#ax.add_patch(r)

dt = 0.016 #計測間隔
epsi = 0.
#Beta = 1000.

# a = MK
a1 = 0.
a2 = 0.
a3 = 0.
a4 = 0.
# b = MB
b1 = 0.
b2 = 0.
b3 = 0.
b4 = 0.
# c = M
c1 = 0.
c2 = 0.
c3 = 0.
c4 = 0.
#Hの中身をかく
h1 = 1.
h2 = 0.
h3 = 0.
h4 = 1.

x = np.array([
    [120.],
    [120.],
    [0.],
    [0.],
    [0.],
    [0.]
]) # 初期位置と初期速度を代入した「6次元状態」 6*1
      
G = np.array([
    [1, 0., 0., 0., 0., 0.],
    [0., 1, 0., 0., 0., 0.],
    [0., 0., 1, 0., 0., 0.],
    [0., 0., 0., 1, 0., 0.],
    [0., 0., 0., 0., 1, 0.],
    [0., 0., 0., 0., 0., 1]
]) #  6*6

Q = np.array([
    [0.00001, 0., 0.,  0., 0., 0.],
    [0., 0.00001, 0.,  0., 0., 0.],
    [0., 0., 0.00001,  0., 0., 0.],
    [0., 0., 0.,  0.00001, 0., 0.],
    [0., 0., 0.,  0., 0.00001, 0.],
    [0., 0., 0.,  0., 0., 0.00001]
]) #ノイズ 6*6

H = np.array([
    [h1, h2, 0., 0., 0., 0.],
    [h3, h4, 0., 0., 0., 0.]
])  # 観測行列 2*6

W = np.array([
    [0.01, 0., 0.,  0., 0., 0.],
    [0., 0.01, 0.,  0., 0., 0.],
    [0., 0., 0.01,  0., 0., 0.],
    [0., 0., 0.,  0.01, 0., 0.],
    [0., 0., 0.,  0., 0.01, 0.],
    [0., 0., 0.,  0., 0., 0.01]
]) #ノイズ 6*6

R = np.array([
    [0.00001, 0.],
    [0., 0.00001]
]) #ノイズ 2*2

#P0 = np.array([
#    [1., 0., 0., 0., 0., 0.],
#    [0., 1.,  0., 0., 0., 0.],
#    [0., 0., 1.,  0., 0., 0.],
#    [0., 0., 0., 1.,  0., 0.],
#    [0., 0., 0., 0., 1.,  0.],
#    [0., 0., 0., 0., 0., 1.]
#]) # 共分散行列 6*6

I = np.identity((len(x)))    # 6次元単位行列
#I = np.identity(6)

mu_D = np.array([
    [0],
    [0],
    [0],
    [0],
    [0],
    [0]
]) #ノイズ 3*1

mu_O = np.array([
    [0],
    [0]
]) #ノイズ 2*1

#global a1,a2,a3,a4,b1,b2,b3,b4,c1,c2,c3,c4
#global E
#global a1,a2,a3,a4,b1,b2,b3,b4,c1,c2,c3,c4
#E=0
IMdKd = 0
IMdBd = 0
IMd = 0
IIMd = 0
Md = 0
Bd = 0
Kd = 0
#beta = 0
#P0 = P

def kalman_filter():
    global a1,a2,a3,a4,b1,b2,b3,b4,c1,c2,c3,c4
    global E
    global IMdKd , IMdBd , IMd, IIMd, Md, Bd, Kd
    global i
    global P0
    global beta
    global j
    global P
    global K
    global S
    global e
  
    x = np.array([
        [points[0][0]],
        [points[0][1]],
        [0.],
        [0.],
        [0.],
        [0.]
    ])

y = np.array([
        [points[0][0]],
        [points[0][1]]
    ])

for n in range(len(points)):

if n == 0:  #0番目は問題なし
            ud = np.array([
                [points[n+1][0]],
                [points[n+1][1]],
                [0],
                [0],
                [0],
                [0],
                [1.],
                [1.]
            ])
            #ud = np.array([
            #    [points[n+1][0]],
            #    [points[n+1][1]],
            #    [points[n+1][2]],
            #    [points[n+1][3]],
            #    [points[n+1][4]],
            #    [points[n+1][5]],
            #    [1.],
            #    [1.]
            #])
            # 外部要素 8*1
        elif n == len(points) - 1: 
            ud = np.array([
                [points[n][0]],
                [points[n][1]],
                [0],
                [0],
                [0],
                [0],
                [1.],
                [1.]
            ])
            #ud = np.array([
            #    [points[n][0]],
            #    [points[n][1]],
            #    [points[n][2]],
            #    [points[n][3]],
            #    [points[n][4]],
            #    [points[n][5]],
            #    [1.],
            #    [1.]
            #])
        else:
            ud = np.array([
                [points[n+1][0]],
                [points[n+1][1]],
                [0],
                [0],
                [0],
                [0],
                [1.],
                [1.]
            ])
            #ud = np.array([
            #    [points[n+1][0]],
            #    [points[n+1][1]],
            #    [points[n+1][2]],
            #    [points[n+1][3]],
            #    [points[n+1][4]],
            #    [points[n+1][5]],
            #    [1.],
            #    [1.]
            #])

xp = float(x[0])
        yp = float(x[1])
        xv = float(x[2])
        yv = float(x[3])
        xa = float(x[4])
        ya = float(x[5])

xpd = float(ud[0])
        ypd = float(ud[1])
        xvd = float(ud[2])
        yvd = float(ud[3])
        xad = float(ud[4])
        yad = float(ud[5])
        Fd = float(ud[6])
        Fi = float(ud[7])

pnx = float(points[n][0])
        pny = float(points[n][1])

Ea1 = -1.0*h1*(-((dt**2)*xp)/2) + (((dt**2)*xpd)/2)*(-h1*(((Fd*c1*(dt**2))/2) + ((Fi*c2*(dt**2))/2) + ((a1*(dt**2)*xpd)/2) - ((a2*(dt**2)*yp)/2) + ((a2*(dt**2)*ypd)/2) + ((b1*(dt**2)*xvd)/2)\
                    - ((b2*(dt**2)*yv)/2) + ((b2*(dt**2)*yvd)/2) + (((dt**2)*xad)/2) + xp*(-((a1*dt**2)/2)+1) + xv*(-((b1*dt**2)/2) + dt))\
                    - h2*(((Fd*c3*dt**2)/2) + ((Fi*c4*dt**2)/2) - ((a3*(dt**2)*xp)/2) + ((a3*(dt**2)*xpd)/2) + ((a4*(dt**2)*ypd)/2) - ((b3*(dt**2)*xv)/2) + ((b3*(dt**2)*xvd)/2) + ((b4*(dt**2)*yvd)/2)\
                    + (((dt**2)*yad)/2) + yp*(-((a4*(dt**2))/2)+1) + yv*(-((b4*dt**2)/2)+dt)) + pnx) - 1.0*h3*(-((dt**2)*xp)/2 + ((dt**2)*xpd)/2)*(-h3*(((Fd*c1*dt**2)/2) + ((Fi*c2*dt**2)/2) + ((a1*(dt**2)*xpd)/2) - ((a2*(dt**2)*yp)/2) + ((a2*(dt**2)*ypd)/2) + ((b1*(dt**2)*xvd)/2)\
                    - ((b2*(dt**2)*yv)/2) + ((b2*(dt**2)*yvd)/2) + (((dt**2)*xad)/2) + xp*(-((a1*(dt**2))/2)+1) + xv*(-((b1*(dt**2))/2)+dt)) - h4*(((Fd*c3*dt**2)/2) + ((Fi*c4*dt**2)/2) - ((a3*(dt**2)*xp)/2) + ((a3*(dt**2)*xpd)/2) + ((a4*(dt**2)*ypd)/2) - ((b3*(dt**2)*xv)/2)\
                    + ((b3*(dt**2)*xvd)/2) + ((b4*(dt**2)*yvd)/2) + (((dt**2)*yad)/2) + yp*(-(a4*(dt**2)/2)+1) + yv*(-(b4*(dt**2)/2)+dt)) + pny)

Ea2 = -1.0*h1*(-((dt**2)*yp)/2) + (((dt**2)*ypd)/2

(Z.T.shape)\n \n e = Z - np.dot(H, x) # 2*1\n\n E += math.sqrt*3\n \n #Sは常にプラス\n S = R + np.dot(np.dot(H, P), H.T) # 2*2\n \n #Sの逆行列も問題なし\n K = np.dot(np.dot(P, H.T), np.linalg.inv(S)) # 6*2\n \n x = x + np.dot(K, e)\n \n P = np.dot*4, P) # 6*6\n\n if n == len(points) - 1 and j == 0 and i == 0:\n P0 = P\n\n beta = beta + (np.linalg.det(P0)/np.linalg.det(P))\n \n #if i == 9 and j == 9:\n #plt.plot(points[n][0],points[n][1],marker='>')\n #plt.plot(x[0], x[1],marker='+')\n #print(y[0],y[1])\n #plt.plot(y[0], y[1],marker='+')\n \n return \n\nfor i in range(10):\n #with codecs.open("simi"+str(i+1)+".csv","r","utf-8","ignore") as file:\n with codecs.open("kawa"+str(i+1)+".csv","r","utf-8","ignore") as file:\n reader = list(csv.reader(file))\n points = [[float(s) for s in z] for z in reader]\n \n P = np.array([\n [1., 0., 0., 0., 0., 0.],\n [0., 1., 0., 0., 0., 0.],\n [0., 0., 1., 0., 0., 0.],\n [0., 0., 0., 1., 0., 0.],\n [0., 0., 0., 0., 1., 0.],\n [0., 0., 0., 0., 0., 1.]\n ]) # 共分散行列 6*6\n\n beta = 0\n \n for j in range(10):\n E = 0\n \n kalman_filter()\n \n #print("結果"+str(i+1)+"回目の誤差", E/len(points)) \n print("結果"+str(i+1)+"回目の信頼度",beta / len(points))\n)

研究会

専門ゼミ

卒業論文テーマ

論文候補

中間発表