Missing behavior data in longitudinal network studies: the impact of treatment methods on estimated effect parameters in stochastic actor oriented models.
Tjeerd ZandbergMark HuismanPublished in: Soc. Netw. Anal. Min. (2019)
Keyphrases
- incomplete data
- statistical methods
- data sets
- learning models
- missing values
- network structure
- experimental data
- data mining techniques
- intrinsic characteristics
- empirical studies
- models built
- prior knowledge
- machine learning methods
- easily interpretable
- data representations
- parameter estimates
- statistical models
- parameter settings
- autoregressive moving average
- data transfer
- predictive model
- bayesian methods
- data mining methods
- input data
- maximum likelihood
- parameter estimation
- database
- spectral clustering
- measured data
- probability distribution
- behavioral data
- high dimensional data
- accurate models
- missing data
- artificial neural networks
- data analysis
- historical data
- decision trees
- training data
- classification models
- data points