Nonlinear Network Reconstruction and Clustering Based on DCOL (Distance Based on Conditional Ordered List) (nlnet)
It includes three methods: K-profiles clustering, non-linear network reconstruction, and non-linear hierarchical clustering.
Statistical Rank Aggregation: Inference, Evaluation, and Visualization (StatRank)
A set of methods to implement Generalized Method of Moments and Maximal Likelihood methods for Random Utility Models. These methods are meant to provide inference on rank comparison data. These methods accept full, partial, and pairwise rankings, and provides methods to break down full or partial rankings into their pairwise components. Please see Generalized Method-of-Moments for Rank Aggregation from NIPS 2013 for a description of some of our methods.
Fits a Fay Herriot Model (smallarea)
Inference techniques for Fay Herriot Model.
Distance Measure Based Judgment and Learning (DJL)
Implements various decision support tools related to the new product development. Subroutines include productivity evaluation using distance measures, benchmarking, risk analysis, technology adoption model, inverse optimization, etc.
Linearized Bregman Algorithms for Generalized Linear Models (Libra)
Efficient procedures for fitting the lasso regularization path for linear regression, logistic and multinomial regression. The package uses Linearized Bregman Algorithm to solve the regularization path through iterations.