BitNet: Bit-Regularized Deep Neural Networks

@article{Raghavan2017BitNetBD,
  title={BitNet: Bit-Regularized Deep Neural Networks},
  author={Aswin Raghavan and Mohamed R. Amer and Sek M. Chai},
  journal={CoRR},
  year={2017},
  volume={abs/1708.04788}
}
We present a novel optimization strategy for training neural networks which we call “BitNet”. The parameters of neural networks are usually unconstrained and have a dynamic range dispersed over all real values. Our key idea is to limit the expressive power of the network by dynamically controlling the range and set of values that the parameters can take. We formulate this idea using a novel end-to-end approach that circumvents the discrete parameter space by optimizing a relaxed continuous and… CONTINUE READING
Recent Discussions
This paper has been referenced on Twitter 32 times over the past 90 days. VIEW TWEETS
1 Citations
49 References
Similar Papers

Citations

Publications citing this paper.

References

Publications referenced by this paper.
Showing 1-10 of 49 references

Similar Papers

Loading similar papers…