-
From Xception to NEXcepTion: New Design Decisions and Neural Architecture Search
Authors:
Hadar Shavit,
Filip Jatelnicki,
Pol Mor-Puigventós,
Wojtek Kowalczyk
Abstract:
In this paper, we present a modified Xception architecture, the NEXcepTion network. Our network has significantly better performance than the original Xception, achieving top-1 accuracy of 81.5% on the ImageNet validation dataset (an improvement of 2.5%) as well as a 28% higher throughput. Another variant of our model, NEXcepTion-TP, reaches 81.8% top-1 accuracy, similar to ConvNeXt (82.1%), while…
▽ More
In this paper, we present a modified Xception architecture, the NEXcepTion network. Our network has significantly better performance than the original Xception, achieving top-1 accuracy of 81.5% on the ImageNet validation dataset (an improvement of 2.5%) as well as a 28% higher throughput. Another variant of our model, NEXcepTion-TP, reaches 81.8% top-1 accuracy, similar to ConvNeXt (82.1%), while having a 27% higher throughput. Our model is the result of applying improved training procedures and new design decisions combined with an application of Neural Architecture Search (NAS) on a smaller dataset. These findings call for revisiting older architectures and reassessing their potential when combined with the latest enhancements.
△ Less
Submitted 28 December, 2022; v1 submitted 16 December, 2022;
originally announced December 2022.
-
Cluster-based Kriging Approximation Algorithms for Complexity Reduction
Authors:
Bas van Stein,
Hao Wang,
Wojtek Kowalczyk,
Michael Emmerich,
Thomas Bäck
Abstract:
Kriging or Gaussian Process Regression is applied in many fields as a non-linear regression model as well as a surrogate model in the field of evolutionary computation. However, the computational and space complexity of Kriging, that is cubic and quadratic in the number of data points respectively, becomes a major bottleneck with more and more data available nowadays. In this paper, we propose a g…
▽ More
Kriging or Gaussian Process Regression is applied in many fields as a non-linear regression model as well as a surrogate model in the field of evolutionary computation. However, the computational and space complexity of Kriging, that is cubic and quadratic in the number of data points respectively, becomes a major bottleneck with more and more data available nowadays. In this paper, we propose a general methodology for the complexity reduction, called cluster Kriging, where the whole data set is partitioned into smaller clusters and multiple Kriging models are built on top of them. In addition, four Kriging approximation algorithms are proposed as candidate algorithms within the new framework. Each of these algorithms can be applied to much larger data sets while maintaining the advantages and power of Kriging. The proposed algorithms are explained in detail and compared empirically against a broad set of existing state-of-the-art Kriging approximation methods on a well-defined testing framework. According to the empirical study, the proposed algorithms consistently outperform the existing algorithms. Moreover, some practical suggestions are provided for using the proposed algorithms.
△ Less
Submitted 4 February, 2017;
originally announced February 2017.
-
DNA Translocation through Graphene Nanopores
Authors:
Grégory F. Schneider,
Stefan W. Kowalczyk,
Victor E. Calado,
Grégory Pandraud,
Henny W. Zandbergen,
Lieven M. K. Vandersypen,
Cees Dekker
Abstract:
Nanopores -- nanosized holes that can transport ions and molecules -- are very promising devices for genomic screening, in particular DNA sequencing. Both solid-state and biological pores suffer from the drawback, however, that the channel constituting the pore is long, viz. 10-100 times the distance between two bases in a DNA molecule (0.5 nm for single-stranded DNA). Here, we demonstrate that it…
▽ More
Nanopores -- nanosized holes that can transport ions and molecules -- are very promising devices for genomic screening, in particular DNA sequencing. Both solid-state and biological pores suffer from the drawback, however, that the channel constituting the pore is long, viz. 10-100 times the distance between two bases in a DNA molecule (0.5 nm for single-stranded DNA). Here, we demonstrate that it is possible to realize and use ultrathin nanopores fabricated in graphene monolayers for single-molecule DNA translocation. The pores are obtained by placing a graphene flake over a microsize hole in a silicon nitride membrane and drilling a nanosize hole in the graphene using an electron beam. As individual DNA molecules translocate through the pore, characteristic temporary conductance changes are observed in the ionic current through the nanopore, setting the stage for future genomic screening.
△ Less
Submitted 26 May, 2010;
originally announced May 2010.
