• Publications
  • Influence
Federated Learning with Non-IID Data
TLDR
We show that the accuracy of federated learning reduces significantly, by up to 55% for neural networks trained for highly skewed non-IID data, where each client device trains on a single class of data. Expand
  • 323
  • 40
  • PDF
CommanderSong: A Systematic Approach for Practical Adversarial Voice Recognition
TLDR
The popularity of ASR (automatic speech recognition) systems, like Google Voice, Cortana, brings in security concerns, as demonstrated by recent attacks. Expand
  • 133
  • 19
  • PDF
Yinyang K-Means: A Drop-In Replacement of the Classic K-Means with Consistent Speedup
TLDR
This paper presents Yinyang K-means, a new algorithm for clustering. Expand
  • 86
  • 17
  • PDF
Planar Graphs of Maximum Degree Seven are Class I
TLDR
In 1965, Vizing proved that planar graphs of maximum degree at least eight have the edge chromatic number equal to their maximum degree. Expand
  • 150
  • 14
  • PDF
PyOD: A Python Toolbox for Scalable Outlier Detection
TLDR
PyOD is an open-source Python toolbox for performing scalable outlier detection on multivariate data, including established outlier ensembles and more recent neural network-based approaches. Expand
  • 96
  • 12
  • PDF
A space-vector modulated sensorless direct-torque control for direct-drive PMSG wind turbines
TLDR
This paper proposes a space-vector modulation (SVM)-based direct-torque control (DTC) scheme for a permanent magnet synchronous generator (PMSG) used in a variable-speed direct-drive wind power generation system. Expand
  • 110
  • 9
DLFuzz: differential fuzzing testing of deep learning systems
TLDR
In this paper, we propose DLFuzz, the first differential fuzzing testing framework to guide DL systems exposing incorrect behaviors. Expand
  • 67
  • 8
  • PDF
The optimal priority models of the intuitionistic fuzzy preference relation and their application in selecting industries with higher meteorological sensitivity
TLDR
This paper proposes a least squares method and a goal programming method to get the priority of the intuitionistic fuzzy preference relation (IFPR) by splitting the IFPR into the membership degree interval judgment matrix and the non-membership degree intervaljudgment matrix. Expand
  • 101
  • 6
  • PDF
A Neighborhood-Based Clustering Algorithm
TLDR
We present a new clustering algorithm, NBC, i.e., Neighborhood Based Clustering, which discovers clusters based on neighborhood characteristics of data. Expand
  • 63
  • 6
  • PDF
A note on the three color problem
TLDR
It is shown that a planar graph withouti-circuits, 4 ≤i ≤ 9, is 3-colorable. Expand
  • 62
  • 6
...
1
2
3
4
5
...