skip to main content

Attention:

The NSF Public Access Repository (NSF-PAR) system and access will be unavailable from 11:00 PM ET on Friday, March 22 until 6:00 AM ET on Saturday, March 23 due to maintenance. We apologize for the inconvenience.


Title: Contextual Dropout: An Efficient Sample-Dependent Dropout Module
Award ID(s):
1812699 1952193
NSF-PAR ID:
10273825
Author(s) / Creator(s):
; ; ; ;
Date Published:
Journal Name:
International Conference on Learning Representations
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. null (Ed.)
  2. null (Ed.)
  3. To mitigate large voltage droop caused by sub-ns dynamic current transitions in system on chips (SoCs), this paper proposes a fully integrated analog-assisted inverter-based digital low dropout regulator (LDO) to obtain a fast response time with 160mV droop at 25mA/100ps featuring 99.4% current efficiency, and 16mV DC load regulation in sub-1V operating range by using a dynamic-step quantizer and a trip-point controller. The proposed quantizer is implemented with an inverter-based flash ADC to achieve high speed without consuming large power while the trip-point controller corrects the DC error of the inverter-based ADC. Besides, the assistant analog LDO is employed to provide fine-grain regulation and remove ripple from the output voltage. 
    more » « less
  4. As severe dropout in single-cell RNA sequencing (scRNA-seq) degrades data quality, current methods for network inference face increased uncertainty from such data. To examine how dropout influences directional dependency inference from scRNA-seq data, we thus studied four methods based on discrete data that are model-free without parametric model assumptions. They include two established methods: conditional entropy and Kruskal-Wallis test, and two recent methods: causal inference by stochastic complexity and function index. We also included three non-directional methods for a contrast. On simulated data, function index performed most favorably at varying dropout rates, sample sizes, and discrete levels. On an scRNA-seq dataset from developing mouse cerebella, function index and Kruskal-Wallis test performed favorably over other methods in detecting expression of developmental genes as a function of time. Overall among the four methods, function index is most resistant to dropout for both directional and dependency inference. The next best choice, Kruskal-Wallis test, carries a directional bias towards a uniformly distributed variable. We conclude that a method robust to marginal distributions with a sufficiently large sample size can reap benefits of single-cell over bulk RNA sequencing in understanding molecular mechanisms at the cellular resolution. 
    more » « less