We introduce equi-tuning, a novel fine-tuning method that transforms (potentially non-equivariant) pretrained models into group equivariant while incurring minimum L_2 loss between the feature representations of and models. Large can be equi-tuned for different groups to satisfy needs various downstream tasks. Equi-tuned benefit from both equivariance as an inductive bias semantic priors provid...

In the Summer of 2008 we have upgraded STAMP to version 8.10 which is the current release version. The new items in version 8.10 are relatively small. The forecasting dialog allows the forecasting of the unobserved components as well as the future observations. More importantly, various errors have been removed from the program and some improvements have been introduced. Most notably, the batch...

Diversity among the pathogenic strains of Theileria equi (T. equi), a major agent of equine piroplasmosis, can affect the appropriate detection of parasite and host immunization. Production of recombinant surface proteins from an infected horse in natural endemic area provides a reliable tool for immunodiagnosis of parasite. Regarding this, the present study was targeted toward the cloning, exp...

Today, neuroscientists are interested in discovering human brain functions through brain networks. In this regard, the evaluation of dynamic changes in functional connectivity of the brain regions by using functional magnetic resonance imaging data has attracted their attention. In this paper, we focus on two model-based approaches, called the exponential weighted moving average model and the d...