Research Article
Biomedical Text Categorization Based on Ensemble Pruning and Optimized Topic Modelling
Table 5
Classification accuracies obtained with different LDA-based configurations.
| ā | Naive Bayes (NB) | Support Vector Machines (SVM) |
| Configuration | oh5 | oh10 | oh15 | ohscal | Ohsu-med | oh5 | oh10 | oh15 | ohscal | Ohsu-med |
| LDA (k=50) | 74.38 | 66.66 | 69.40 | 59.27 | 28.35 | 76.24 | 78.73 | 83.17 | 70.62 | 34.64 | LDA (k=100) | 70.85 | 63.64 | 67.44 | 60.05 | 29.56 | 78.28 | 78.25 | 83.23 | 73.23 | 38.82 | LDA (k=150) | 69.02 | 65.24 | 65.51 | 59.01 | 29.43 | 76.72 | 79.09 | 84.74 | 73.8 | 41.27 | LDA (k=200) | 66.17 | 64.01 | 63.61 | 58.93 | 27.99 | 77.33 | 77.93 | 84 | 74.19 | 41.82 | GA-LDA (BIC) | 75.16 | 67.24 | 74.70 | 71.66 | 35.45 | 77.98 | 69.03 | 75.12 | 73.62 | 35.83 | PSO-LDA (BIC) | 75.40 | 68.60 | 76.90 | 72.43 | 35.46 | 78.22 | 72.56 | 75.17 | 75.89 | 36.23 | FA-LDA (BIC) | 75.48 | 71.26 | 77.48 | 72.80 | 35.60 | 79.50 | 74.73 | 76.63 | 76.90 | 37.69 | CSA-LDA (BIC) | 76.66 | 71.96 | 78.77 | 72.94 | 35.65 | 79.56 | 75.97 | 77.96 | 77.02 | 37.94 | BA-LDA (BIC) | 78.82 | 72.21 | 79.77 | 73.02 | 36.58 | 79.85 | 76.53 | 78.89 | 77.34 | 38.89 | GA-LDA (CH) | 79.02 | 72.88 | 80.11 | 74.53 | 36.85 | 80.62 | 77.72 | 80.31 | 78.17 | 38.96 | PSO-LDA (CH) | 80.20 | 72.93 | 80.66 | 74.76 | 37.03 | 81.50 | 77.91 | 80.50 | 78.99 | 39.03 | FA-LDA (CH) | 81.20 | 72.99 | 80.72 | 75.13 | 37.75 | 81.80 | 77.99 | 80.55 | 79.09 | 39.03 | CSA-LDA (CH) | 81.40 | 73.12 | 81.71 | 76.02 | 38.34 | 82.61 | 78.01 | 80.78 | 79.82 | 39.03 | BA-LDA (CH) | 81.46 | 73.49 | 81.82 | 76.21 | 39.24 | 82.87 | 78.93 | 81.01 | 79.89 | 39.52 | GA-LDA (DB) | 84.46 | 76.22 | 84.13 | 78.71 | 40.50 | 84.73 | 80.95 | 85.88 | 82.46 | 43.02 | PSO-LDA (DB) | 84.60 | 80.07 | 85.14 | 79.21 | 42.57 | 85.13 | 81.11 | 86.17 | 84.22 | 43.51 | FA-LDA (DB) | 85.89 | 80.82 | 85.17 | 80.83 | 44.60 | 86.22 | 81.88 | 86.73 | 84.62 | 44.61 | CSA-LDA (DB) | 86.42 | 80.97 | 86.10 | 81.69 | 45.21 | 86.79 | 82.00 | 86.96 | 85.07 | 46.67 | BA-LDA (DB) | 87.60 | 81.36 | 87.32 | 83.56 | 47.00 | 88.86 | 82.09 | 88.05 | 85.24 | 50.08 | GA-LDA (SI) | 81.57 | 73.57 | 82.03 | 76.48 | 39.36 | 83.21 | 79.00 | 82.24 | 79.93 | 40.58 | PSO-LDA (SI) | 82.61 | 73.76 | 82.50 | 76.61 | 39.66 | 83.58 | 79.33 | 83.03 | 80.36 | 40.87 | FA-LDA (SI) | 83.19 | 74.18 | 82.88 | 77.47 | 39.68 | 83.69 | 79.41 | 83.11 | 80.95 | 40.95 | CSA-LDA (SI) | 83.78 | 75.11 | 83.01 | 78.06 | 39.69 | 83.84 | 80.83 | 84.47 | 81.82 | 41.12 | BA-LDA (SI) | 84.11 | 76.08 | 83.03 | 78.13 | 40.08 | 84.49 | 80.90 | 85.52 | 81.99 | 42.65 |
|
|
LDA: latent Dirichlet allocation, GA-LDA: genetic algorithm based LDA, PSO-LDA: particle swarm optimization based LDA, FA-LDA: firefly algorithm based LDA, CSA-LDA: cuckoo search algorithm based LDA, BA-LDA: bat algorithm based LDA, BIC: Bayesian information criterion, CH: Calinski-Harabasz index, DB: Davies-Bouldin index, and SI: Silhouette index.
|