Bridging the Gap Between Graph Edit Distance and Kernel by Michel Neuhaus

By Michel Neuhaus

In graph-based structural trend acceptance, the belief is to remodel styles into graphs and practice the research and popularity of styles within the graph area - generally often called graph matching. lots of equipment for graph matching were proposed. Graph edit distance, for example, defines the dissimilarity of 2 graphs through the quantity of distortion that's had to remodel one graph into the opposite and is taken into account probably the most versatile equipment for error-tolerant graph matching.This e-book makes a speciality of graph kernel services which are hugely tolerant in the direction of structural mistakes. the fundamental thought is to include thoughts from graph edit distance into kernel features, therefore combining the pliability of edit distance-based graph matching with the facility of kernel machines for trend popularity. The authors introduce a suite of novel graph kernels on the topic of edit distance, together with diffusion kernels, convolution kernels, and random stroll kernels. From an experimental assessment of a semi-artificial line drawing info set and 4 real-world facts units inclusive of photos, microscopic photographs, fingerprints, and molecules, the authors show that the various kernel features together with help vector machines considerably outperform conventional edit distance-based nearest-neighbor classifiers, either when it comes to class accuracy and operating time.

Show description

Read Online or Download Bridging the Gap Between Graph Edit Distance and Kernel Machines (Series in Machine Perception and Artificial Intelligence) PDF

Similar computer vision & pattern recognition books

Kernel methods in computer vision

Few advancements have prompted the sector of computing device imaginative and prescient within the final decade greater than the advent of statistical laptop studying concepts. rather kernel-based classifiers, reminiscent of the aid vector laptop, became crucial instruments, delivering a unified framework for fixing a variety of image-related prediction initiatives, together with face popularity, item detection, and motion type.

Fourier Vision: Segmentation and Velocity Measurement using the Fourier Transform

Fourier imaginative and prescient offers a brand new therapy of figure-ground segmentation in scenes comprising obvious, translucent, or opaque gadgets. Exploiting the relative movement among determine and floor, this method bargains explicitly with the separation of additive signs and makes no assumptions concerning the spatial or spectral content material of the photographs, with segmentation being conducted phasor via phasor within the Fourier area.

Two-Dimensional Change Detection Methods: Remote Sensing Applications

Switch detection utilizing remotely sensed pictures has many purposes, corresponding to city tracking, land-cover switch research, and catastrophe administration. This paintings investigates two-dimensional switch detection tools. the prevailing tools within the literature are grouped into 4 different types: pixel-based, transformation-based, texture analysis-based, and structure-based.

Evolution of Broadcast Content Distribution

This booklet discusses possibilities for broadcasters that come up with the appearance of broadband networks, either mounted and cellular. It discusses how the conventional manner of dispensing audio-visual content material over broadcasting networks has been complemented by means of the use of broadband networks. the writer exhibits how this additionally provides the chance to provide new different types of interactive or so-called nonlinear prone.

Additional resources for Bridging the Gap Between Graph Edit Distance and Kernel Machines (Series in Machine Perception and Artificial Intelligence)

Example text

The optimal area of the diagram is obivously in the upper left corner, which corresponds to fast and simultaneously very accurate systems. Note that the running time of the exact distance is much too high to be included in the illustration. It takes only 4 minutes for the fastest approximate classifier among those with maximum accuracy to classify the 1,600 fingerprint graphs, whereas the running time of the exact algorithm amounts to over 19 hours for the same task. Yet, in terms of classification accuracy, the exact method outperforms all approximate systems.

A single discrete attribute is attached to each edge representing its direction. 1 Running time and resulting distances of exact edit distance algorithm (1 run) and approximate edit distance algorithm (50 runs) Nodes 5 7 9 12 20 30 42 169 a Running time Exact Approx. Edit distance Exact Approx. 5 Empty entries indicate failure due to lack of memory. responding graph, see Fig. 6. The problem of fingerprint classification is treated in greater detail in Sec. 2. As mentioned previously, graph edit distance is usually only applicable to graphs containing up to around 12 nodes, but the extracted graphs typically consist of about 170 nodes and 190 edges.

In Fig. 2, the node substitutions of the optimal edit path between the two graphs are visualized. If two graphs are even less similar, such as those in Fig. 3, the edit distance is still capable of identifying how one graph is mapped onto the other graph in an optimal way. Note that in these illustrations thin lines connecting two gray nodes represent node substitutions, and white nodes represent nodes to be deleted or inserted. July 21, 2007 21:8 World Scientific Book - 9in x 6in Graph Edit Distance Fig.

Download PDF sample

Rated 4.28 of 5 – based on 22 votes