Tensor-variate Mixture of Experts for Proportional Myographic Control of a Robotic Hand
This work addresses overfitting in tensor regression for robotic hand control, offering a domain-specific improvement for myographic applications.
The paper tackled the problem of regression with tensor-valued data by proposing a mixture-of-experts model that preserves data structure to avoid overfitting with limited training data, achieving effectiveness in hand movement recognition from tactile myography in offline and real-time experiments.
When data are organized in matrices or arrays of higher dimensions (tensors), classical regression methods first transform these data into vectors, therefore ignoring the underlying structure of the data and increasing the dimensionality of the problem. This flattening operation typically leads to overfitting when only few training data is available. In this paper, we present a mixture-of-experts model that exploits tensorial representations for regression of tensor-valued data. The proposed formulation takes into account the underlying structure of the data and remains efficient when few training data are available. Evaluation on artificially generated data, as well as offline and real-time experiments recognizing hand movements from tactile myography prove the effectiveness of the proposed approach.