GriMa: A Grid Mining Algorithm for Bag-of-Grid-Based Classification

Abstract

Abstract. General-purpose exhaustive graph mining algorithms have seldom been used in real life contexts due to the high complexity of the process that is mostly based on costly isomorphism tests and countless expansion possibilities. In this paper, we explain how to exploit gridbased representations of problems to efficiently extract frequent grid subgraphs and create Bag-of-Grids which can be used as new features for classification purposes. We provide an efficient grid mining algorithm called GriMA which is designed to scale to large amount of data. We apply our algorithm on image classification problems where typical Bagof-Visual-Words-based techniques are used. However, those techniques make use of limited spatial information in the image which could be beneficial to obtain more discriminative features. Experiments on different datasets show that our algorithm is efficient and that adding the structure may greatly help the image classification process.

Mining Frequent Patterns in 2D+t Grid Graphs for Cellular Automata Analysis

“…This corresponds to a "node-induced" closure operator on graphs where, given a pattern P , we add all possible edges to P without adding new nodes. We have shown in [6] that this optimization decreases the number of extracted patterns and the extraction time.…”

“…Therefore, FreqGeo may be used to mine grids. However, it has a higher time-complexity than GriMA, the 2D grid mining algorithm introduced in [6].…”

“…A 2D grid is a particular geometric graph that may be used to model any 2D regular structure such as, for example, grids of pixels (i.e., images), game boards, or cellular automata. To characterize these grids, we may mine them to extract recurrent patterns [6]. In some applications, we use temporal sequences of grids (i.e., 2D+t grids) to model the evolution of grids through time.…”

“…Our algorithm is an extension of GriMA [6], an algorithm for mining 2D grids which has been designed to tackle real-life applications such as image classification. This algorithm is recalled in Section 2.…”