| Figure 15.1 - LSI and the SVD.png | 32.2 KB | |
| Figure 15.10 - Most common earnings call toke ns.png | 90.5 KB | |
| Figure 15.11 - Most important words for earnings call topics.png | 261.2 KB | |
| Figure 15.12 - pyLDAVis' interactive topic explorer.png | 253.8 KB | |
| Figure 15.13 - Impact of LDA hyperparameter settings on topic quality.png | 253.9 KB | |
| Figure 15.14 - Corpus statistics for financial news data.png | 102.6 KB | |
| Figure 15.15 - Top 15 words for financial news topics.png | 347.5 KB | |
| Figure 15.2 - LSI topic weights for train and test data.png | 88.1 KB | |
| Figure 15.3 - Top 10 words per LSI topic.png | 190.5 KB | |
| Figure 15.4 - The statistical dependencies modeled by pLSA in plate notation.png | 18.3 KB | |
| Figure 15.5 - pLSA weights by topic for train and test data.png | 101 KB | |
| Figure 15.6 - Top words per topic for p LSA.png | 179.2 KB | |
| Figure 15.7 - Three draws from the Dirichlet distribution.png | 9.6 KB | |
| Figure 15.8 - The statistical dependencies of the LDA model in plate notation.png | 46 KB | |
| Figure 15.9 - Topic coherence and test set assignments.png | 74.2 KB | |
