ã¹ãã¹ãã®ãããæ§ã¨ããæ©ç¨®ãããã¾ããBIGé£ãã¡ã¤ã³ã®åºãæ¹ã«ãªãã¾ãã
é常æã¯ãã£ã³ã¹ç®ãå¼ãã¦BIGorREGã®å½é¸ããããã¾ãã
æ©æ¢°å²ããã£ã¨ãä½ãã®ã¯
ãã£ã³ã¹ã¢ã¼ããAã«ãã
æã§ãã
ãããåé¿ããããã«ã¢ã¼ããæ¨æ¸¬ãã¾ã
- ãã£ã³ã¹ç®ã®è¦å®åæ°ã¨ã¯
- ã¢ã¼ãã®åæåã«ã¤ãã¦
- ã¢ã¼ã移è¡ã«ã¤ãã¦
- ã¢ã¼ãæ¨æ¸¬ã«ã¤ãã¦
- ã¹ã«ã¼ã«ã¤ãã¦
ãã£ã³ã¹ç®ã®è¦å®åæ°ã¨ã¯
ã¢ã¼ãã«ãã£ã¦åæ°ã決ãããã¦ããã
ã¡ããã¼ãããæ§ããå¼ç¨
ãã£ã³ã¹ç®ã®è¦å®åæ°ã管çããã¢ã¼ãã4種é¡åå¨ã
ã¢ã¼ããé«ãã»ã©ãã£ã³ã¹ç®ã®è¦å®åæ°æ¯ãåããåªéãããã
ã¢ã¼ãã®åæåã«ã¤ãã¦
è¨å®å¤æ´å¾ã¨BIGå¤æ´å¾ããã
ã¢ã¼ã | ãªã»ãã | BIGå¾ |
---|---|---|
A | 38% | 62% |
B | 60% | 35% |
C | 1% | 2% |
D | 1% | 1% |
ãªã»ããæ©æµã¯å¤©äºã800Gï¼Î±ãã¡ã¤ã³ãªã®ã§ ã¢ã¼ãã«é¢ãã¦ã¯ã¯ããã¨å¬ããã¬ãã«ã«ãªã£ã¦ãã
ã¢ã¼ã移è¡ã«ã¤ãã¦
CZ失ææã»REGçµäºæã«ã¢ã¼ã移è¡ããããªãã
è¦ãã°ããããã¢ã¼ãC以ä¸ã«ç§»è¡ããã°BIGã«å½é¸ããã¾ã§ã¢ã¼ãC以ä¸ã¨ãªãã
ã¢ã¼ãæ¨æ¸¬ã«ã¤ãã¦
CZ失æãREGãçºçããã¨ã¢ã¼ãããããä»æ§ãªã®ã§
ã©ããããã®ç¢ºçã§ã¢ã¼ãã«ããã®ããè¨ç®ãã¦ã¿ãã
BIGã¨ãªã»ããç¶æ
ã§ã¯å¾®å¦ã«éãã®ã§å¥ã
ã«ãã¦ãã¾ã
BIGå¾ã®æ¨æ¸¬ã«ã¤ãã¦
ã¢ã¼ãA | ã¢ã¼ãB | ã¢ã¼ãC | ã¢ã¼ãD | |
---|---|---|---|---|
1åã¹ã«ã¼ | 23.560% | 57.150% | 17.430% | 1.860% |
2åã¹ã«ã¼ | 8.953% | 46.711% | 37.486% | 6.850% |
3åã¹ã«ã¼ | 3.402% | 31.997% | 50.336% | 14.265% |
4åã¹ã«ã¼ | 1.293% | 20.280% | 58.400% | 20.028% |
5åã¹ã«ã¼ | 0.491% | 12.335% | 63.263% | 23.911% |
6åã¹ã«ã¼ | 0.187% | 7.326% | 66.137% | 26.350% |
7åã¹ã«ã¼ | 0.071% | 4.288% | 67.815% | 27.827% |
8åã¹ã«ã¼ | 0.027% | 2.487% | 68.786% | 28.701% |
9åã¹ã«ã¼ | 0.010% | 1.434% | 69.345% | 29.212% |
10åã¹ã«ã¼ | 0.004% | 0.823% | 69.665% | 29.508% |
11åã¹ã«ã¼ | 0.001% | 0.472% | 69.849% | 29.678% |
12åã¹ã«ã¼ | 0.001% | 0.270% | 69.954% | 29.776% |
6ã¹ã«ã¼ããã¯90%以ä¸ã§Cã«ããã
ãªã»ããå¾ã®æ¨æ¸¬ã«ã¤ãã¦
ã¢ã¼ãA | ã¢ã¼ãB | ã¢ã¼ãC | ã¢ã¼ãD | |
---|---|---|---|---|
1åã¹ã«ã¼ | 14.440% | 57.000% | 27.010% | 1.550% |
2åã¹ã«ã¼ | 5.487% | 41.154% | 43.614% | 9.745% |
3åã¹ã«ã¼ | 2.085% | 26.750% | 54.306% | 16.859% |
4åã¹ã«ã¼ | 0.792% | 16.499% | 60.828% | 21.881% |
5åã¹ã«ã¼ | 0.301% | 9.880% | 64.711% | 25.108% |
6åã¹ã«ã¼ | 0.114% | 5.812% | 66.987% | 27.086% |
7åã¹ã«ã¼ | 0.043% | 3.382% | 68.308% | 28.267% |
8åã¹ã«ã¼ | 0.017% | 1.954% | 69.070% | 28.960% |
9åã¹ã«ã¼ | 0.006% | 1.123% | 69.508% | 29.362% |
10åã¹ã«ã¼ | 0.002% | 0.644% | 69.759% | 29.594% |
11åã¹ã«ã¼ | 0.001% | 0.369% | 69.903% | 29.728% |
12åã¹ã«ã¼ | 0.000% | 0.211% | 69.985% | 29.805% |
ãªã»ããå¾ã6ã¹ã«ã¼ããã¯90%以ä¸ã§Cã«ããã 5ã¹ã«ã¼ã§ã89%ã§Cã«ãã
è¨ç®ããã°ã©ã
è¦ã人ã¯ããªãã¨æããã©è¨ç®ã«ã¤ãã£ãããã°ã©ã ã§ã
ã¯ãªãã¯ããã¨å±éããã¾ã
import numpy as np import pandas as pd # BIGå¾ # åæ確ç initial_probs = np.array([0.62, 0.35, 0.02, 0.01]) # é·ç§»è¡å transition_matrix = np.array([ [0.38, 0.60, 0.01, 0.01], # A -> [A, B, C, D] [0, 0.57, 0.42, 0.01], # B -> [A, B, C, D] [0, 0, 0.68, 0.32], # C -> [A, B, C, D] [0, 0, 0.75, 0.25 ] # D -> [A, B, C, D] ]) # 1åãã12åã®ã¹ã«ã¼å¾ã®ç¢ºçãè¨ç® results = {} for i in range(1, 13): results[f"{i}åã¹ã«ã¼"] = initial_probs @ np.linalg.matrix_power(transition_matrix, i) # çµæãDataFrameã«å¤æãããã¼ã»ã³ã表示ã«å¤æ result_df = pd.DataFrame(results, index=['ã¢ã¼ãA', 'ã¢ã¼ãB', 'ã¢ã¼ãC', 'ã¢ã¼ãD']).T result_df = result_df * 100 # ãã¼ã»ã³ãã«å¤æ result_df = result_df.round(3) # å°æ°ç¹3æ¡ã§ä¸¸ãã print(result_df) # åæ確çï¼æä¸ï¼ initial_probs = np.array([0.38, 0.60, 0.01, 0.01]) # é·ç§»è¡åï¼æä¾ããã表ã«åºã¥ãï¼ transition_matrix = np.array([ [0.38, 0.60, 0.01, 0.01], # A -> [A, B, C, D] [0, 0.57, 0.42, 0.01], # B -> [A, B, C, D] [0, 0, 0.68, 0.32], # C -> [A, B, C, D] [0, 0, 0.75, 0.25 ] # D -> [A, B, C, D] ]) # 1åãã10åã®ã¹ã«ã¼å¾ã®ç¢ºçãè¨ç® results = {} for i in range(1, 13): results[f"{i}åã¹ã«ã¼"] = initial_probs @ np.linalg.matrix_power(transition_matrix, i) # çµæãDataFrameã«å¤æãã¦è¡¨ç¤º result_df = pd.DataFrame(results, index=['ã¢ã¼ãA', 'ã¢ã¼ãB', 'ã¢ã¼ãC', 'ã¢ã¼ãD']).T result_df = result_df * 100 # ãã¼ã»ã³ãã«å¤æ result_df = result_df.round(3) # å°æ°ç¹3æ¡ã§ä¸¸ãã print(result_df)
å®éã«åãããã人ã¯colabã§åãããã¨ãã§ãã¾ã
ã¹ã«ã¼ã«ã¤ãã¦
REGã¯ãã¼ã¿ã«ã¦ã³ã¿ã¼ã«ããã£ã¦ããããåé¡ãªãã CZã¯ãããªãï¼ãã¤ãã§ã¯ãããªããããããã¼ã¿ã«ã¦ã³ã¿ã¼ãããï¼ èªåã¯ãé常Aã§ãã£ã¦ã300Gãããã§CZã¯ãããããã¨æ³å®ãã¦ãã