This article proposes a novel similarity measure between
vector sequences. Recently, a model-based approach was
introduced to address this issue. It consists in modeling
each se...
- An inversion algorithm is developed to recover power and duty-cycle of incoming Radio Frequency Interference (RFI) signals from kurtosis. The algorithm applies simulated annealin...
Existing word similarity measures are not robust to data sparseness since they rely only on the point estimation of words' context profiles obtained from a limited amount of ...
Jun'ichi Kazama, Stijn De Saeger, Kow Kuroda, Masa...
A novel approach to measure the interdependence of two time series is proposed, referred to as “stochastic event synchrony” (SES); it quantifies the alignment of two point pr...
We introduce perturbation kernels, a new class of similarity measure for information retrieval that casts word similarity in terms of multi-task learning. Perturbation kernels mode...