Table 2 Time series and supervised learning methods best results (lowest RAE) for three periods (1991–2011, 2002–2011 and 2007–2011) having 2012 as the year of prediction
From: Improving trend analysis using social network features
Period | Time series regression | Time series supervised learning methods |
|---|---|---|
1991–2011 | 113.16% | 51.52% |
2002–2011 | 136.42% | 51.01% |
2007–2011 | 288.14% | 50.31% |