使用 NMF非负矩阵分解

如何解决使用 NMF非负矩阵分解

我正在处理交通数据以估计公交车的到达时间。我正在尝试使用 NMF 将路段（从一站 I 到下一站 I+1 的每个路段）聚类到 k 个集群中，其中每个集群都有基于交通拥堵和流量具有相似旅行时间的路段。主矩阵 X 是 NM，其中 N 是路线中的路段数，M 是一天中的间隔次数，Xi,j 是链接 i 在间隔时间 j 的行程时间。我正在使用 Sklearn.decomposition.NMF 模型来查找 WH 矩阵，如下所示：

from sklearn.decomposition import NMF
model = NMF(n_components=2,init='random',random_state=0)
W = model.fit_transform(X)
H = model.components_

知道如何找到集群的数量以及与每个集群相关的段。

这是我得到的 X、W、H 矩阵：

X = [[186.997151,146.06308411,163.66979167,197.02702703,157.95852535,163.79672131],[156.84571429,154.68997669,141.75208333,138.38121547,144.87557604,137.99673913],[111.14204545,93.59815242,107.81452456,85.76836158,83.40552995,80.79761905],[172.78531073,101.10798122,169.4069828,111.59887006,97.91324201,116.97830803],[117.16524217,98.35915493,109.72216441,97.91620112,88.93150685,93.86413043],[158.76589595,164.89095128,148.14568158,135.24719101,152.43636364,148.10978261],[ 92.60237389,74.80510441,78.36354221,81.4180791,74.6712963,79.96103896],[ 95.9704142,79.44779582,75.36839378,70.30508475,73.18807339,75.14208243],[180.24251497,177.25,141.78278901,124.78735632,145.05909091,134.81818182],[173.25688073,157.72146119,138.53826267,134.27428571,148.12785388,135.21103896],[ 95.05993691,75.17155756,78.2547705,87.29545455,74.46818182,79.88286334],[ 98.92971246,125.66517857,97.4623323,93.61797753,105.53603604,97.34422658],[ 83.83227848,76.25950783,74.10227273,89.42613636,83.54954955,76.88152174],[164.43589744,123.43340858,124.15277778,111.86781609,134.17272727,121.64247021],[262.13694268,227.75900901,207.27099433,204.79289941,225.91441441,206.53613808],[182.62376238,126.23608018,123.28174807,148.26190476,113.37946429,114.68722944],[220.24013158,226.21524664,181.2319138,182.95375723,189.07727273,183.11243243]]
W = [[1.01156082e+01,5.06953223e+00,1.27913100e+00,6.69122207e+00,6.23150300e+00,8.63174749e+00],[8.32873962e+00,5.36311809e+00,4.52776726e+00,6.47717651e+00,4.86835296e+00,3.22449667e+00],[8.69655455e+00,6.43261438e+00,1.98844191e+00,4.41714626e+00,0.00000000e+00,0.00000000e+00],[9.15155374e+00,1.26416808e+01,1.25401841e-01,8.93319149e+00,4.21471503e-01,5.66643983e-01],[6.61013376e+00,6.31173335e+00,2.37508475e+00,4.92230568e+00,1.69539540e+00,1.97750748e+00],[9.75057120e+00,4.94748811e+00,5.16241556e+00,6.99774312e+00,4.51566723e+00,1.94041373e+00],[7.73686890e-02,3.95504771e+00,1.94017427e+00,4.60412327e+00,4.98716865e+00,4.37181112e+00],[3.67364290e-01,4.88288534e+00,2.24252584e+00,4.39207344e+00,4.65811925e+00,2.90560699e+00],[1.03282368e+01,9.18371538e+00,5.56077107e+00,5.64160329e+00,3.57861366e+00,[1.30479796e+01,5.72828693e+00,2.73026281e+00,5.15980486e+00,3.67442603e+00,8.38380012e-01],[3.40533981e-01,4.17491497e+00,1.81767859e+00,4.27043603e+00,4.88918779e+00,4.87104216e+00],[2.80683468e+00,3.92567950e+00,6.31450347e+00,4.89366154e+00,4.06225948e+00,2.60290253e+00],[1.89292479e+00,1.42093084e+00,1.70573943e+00,3.75436103e+00,4.93914176e+00,4.69166906e+00],[8.74271245e+00,4.63318972e+00,1.76838809e-02,6.12468640e+00,6.16663616e+00,1.54101772e+00],[1.99015285e+01,7.16928280e+00,2.52906778e+00,7.96289699e+00,6.47447755e+00,1.86495523e+00],[1.21617439e+01,9.48305232e+00,2.83506384e+00,2.36955986e+00,3.74107478e+00],[1.79947086e+01,8.93629831e+00,6.63479012e+00,5.76440688e+00,2.64375500e+00,9.12967294e-01]]
H = [[ 6.48633465,6.26879206,4.6279997,6.45626077,6.23821414,5.33766419],[ 7.48801066,3.05759946,2.43478263,2.07355664,0.19533404,1.06692284],[ 0.,9.50253736,1.87738386,3.02032145,4.65012356,3.87646987],[ 1.54598381,0.02979102,10.55127615,2.15804718,3.79700524,5.50353502],[ 9.99909359,8.76047554,0.38871719,2.78649831,8.71044715,5.05421432],[ 1.2450878,0.03008991,3.40035723,9.92750215,0.9959392,3.51367207]]