图共 5个 表共 3
    • 图  1  基于变分自编码器的深度嵌入聚类

      Figure 1.  Deep embedded clustering based on the variational auto encoder

    • 图  2  聚类框架中的变分编码器结构

      Figure 2.  Structure of the encoder of variational auto encoder in the clustering framework

    • 图  3  8种编码DNA载体的阻断电流信号

      Figure 3.  Blockade signals of eight encoded DNA carriers

    • 图  4  0 ~ 3类中离聚类中心最近的5个样本

      Figure 4.  Five samples closest to the centroids of cluster 0 to 3

    • 图  5  降采样后的聚类结果

      Figure 5.  Clustering results after under-sampling

    • LabelNumber
      0005 838
      0018 663
      0102 410
      01115 990
      100963
      1017 668
      1107 066
      1117 391

      表 1  编码DNA载体的阻断事件数量

      Table 1.  Numbers of blockade events produced by encoded DNA carriers

    • Clustering methodACC
      K-means0.214 4
      AE + K-means0.220 0
      VAE + K-means0.236 7
      IDEC0.243 0
      VAE + IDEC0.277 5

      表 2  K-means、AE + K-means、VAE + K-means、IDEC、VAE + IDEC的聚类结果比较(8个聚类中心)

      Table 2.  Clustering results comparison of K-means, AE + K-means, VAE + K-means, IDEC, VAE + IDEC (8 clusters)

    • Clustering methodACC
      K-means0.269 0
      AE + K-means0.254 6
      VAE + K-means0.248 0
      IDEC0.253 6
      VAE + IDEC0.305 8

      表 3  K-means、AE + K-means、VAE + K-means、IDEC、VAE + IDEC的聚类结果比较(6个聚类中心)      

      Table 3.  Clustering results comparison of K-means, AE + K-means, VAE + K-means, IDEC, VAE + IDEC (6 clusters)