Skip to main content

Table 2 Overall prediction performance

From: Information cascades prediction with attention neural network

 

t = 1*

t = 2*

t = 3*

MSE

MAE

MSE

MAE

MSE

MAE

Twitter

 Features-linear

3.821

1.536

3.511

1.479

3.423

1.42

 SVR

3.798

1.529

3.028

1.384

3.382

1.411

 SEISMIC

3.770

1.527

2.954

1.313

1.462

3.319

1.408

 DeepCas

3.725

1.493

3.496

 

3.308

1.395

 Proposed

2.609

1.265

2.349

1.167

2.300

1.14

AMiner

 Features-linear

2.429

1.197

2.136

1.089

1.880

1.067

 SVR

2.419

1.194

2.195

1.123

1.865

1.066

 SEISMIC

2.417

1.193

2.282

1.136

1.852

1.061

 DeepCas

2.239

1.127

1.987

1.072

1.674

1.056

 Proposed

2.172

1.116

1.672

1.042

1.534

1.003

  1. p.s. t = 1*, where ‘*’ denotes hour for Twitter (year for AMiner)
Back to article page