NEC

US9870519

Title of Patent:

Hierarchical sparse dictionary learning (HiSDL) for heterogeneous high-dimensional time series

• A hierarchical sparse dictionary learning system and method to construct a learned dictionary regularized by a prior over-complete dictionary

• Possible to characterize given data which includes high-dimensional time series for an interpretable dictionary output which can adapt to given data and generalize by a priori dictionaries

• Able to determine whether the learned dictionary converges to an input data set, and output the learned dictionary regularized by at least one a priori over-complete dictionary

  
  

  
  

NEC

US11423647

Title of Patent:

Identification system, model re-learning method and program

• Identifying an object indicated by data, by applying the data to a model, a model re-learning method and a model re-learning program applied to the identification system

• The computer can execute the learning processing indicated by using training data

• Able to re-learn a model to improve identification accuracy of the model for identifying an object indicated by data

  
  

Potential Areas of Applications:

• Artificial Intelligence / Machine Learning

• Banking and Finance

• Robotics

• Entertainment Streaming

Proposed Schemes:

Assignment basis 8% - 12% equity exchange

* Schemes are subject to the business model and are open for discussion.