An official website of the United States government
Persistence diagrams have been widely used to quantify the underlying features of filtered topological spaces in data visualization. In many applications, computing distances between diagrams is essential; however, computing these distances has been challenging due to the computational cost. In this paper, we propose a persistence diagram hashing framework that learns a binary code representation of persistence diagrams, which allows for fast computation of distances. This framework is built upon a generative adversarial network (GAN) with a diagram distance loss function to steer the learning process. Instead of using standard representations, we hash diagrams into binary codes, which have natural advantages in large-scale tasks. The training of this model is domain-oblivious in that it can be computed purely from synthetic, randomly created diagrams. As a consequence, our proposed method is directly applicable to various datasets without the need for retraining the model. These binary codes, when compared using fast Hamming distance, better maintain topological similarity properties between datasets than other vectorized representations. To evaluate this method, we apply our framework to the problem of diagram clustering and we compare the quality and performance of our approach to the state-of-the-art. In addition, we show the scalability of our approach on a dataset with 10k persistence diagrams, which is not possible with current techniques. Moreover, our experimental results demonstrate that our method is significantly faster with the potential of less memory usage, while retaining comparable or better quality comparisons.
more »
« less
Learning Hyperbolic Representations of Topological Features
Learning task-specific representations of persistence diagrams is an important problem in topological data analysis and machine learning. However, current state of the art methods are restricted in terms of their expressivity as they are focused on Euclidean representations. Persistence diagrams often contain features of infinite persistence (i.e., essential features) and Euclidean spaces shrink their importance relative to non-essential features because they cannot assign infinite distance to finite points. To deal with this issue, we propose a method to learn representations of persistence diagrams on hyperbolic spaces, more specifically on the Poincare ball. By representing features of infinite persistence infinitesimally close to the boundary of the ball, their distance to non-essential features approaches infinity, thereby their relative importance is preserved. This is achieved without utilizing extremely high values for the learnable parameters, thus the representation can be fed into downstream optimization methods and trained efficiently in an end-to-end fashion. We present experimental results on graph and image classification tasks and show that the performance of our method is on par with or exceeds the performance of other state of the art methods.
more »
« less
- Award ID(s):
- 1932620
- PAR ID:
- 10302276
- Date Published:
- Journal Name:
- Proceedings of the International Conference on Learning Representations (ICLR)
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Goaoc, Xavier and (Ed.)The space of persistence diagrams under bottleneck distance is known to have infinite doubling dimension. Because many metric search algorithms and data structures have bounds that depend on the dimension of the search space, the high-dimensionality makes it difficult to analyze and compare asymptotic running times of metric search algorithms on this space. We introduce the notion of nearly-doubling metrics, those that are Gromov-Hausdorff close to metric spaces of bounded doubling dimension and prove that bounded $$k$$-point persistence diagrams are nearly-doubling. This allows us to prove that in some ways, persistence diagrams can be expected to behave like a doubling metric space. We prove our results in great generality, studying a large class of quotient metrics (of which the persistence plane is just one example). We also prove bounds on the dimension of the $$k$$-point bottleneck space over such metrics. The notion of being nearly-doubling in this Gromov-Hausdorff sense is likely of more general interest. Some algorithms that have a dependence on the dimension can be analyzed in terms of the dimension of the nearby metric rather than that of the metric itself. We give a specific example of this phenomenon by analyzing an algorithm to compute metric nets, a useful operation on persistence diagrams.more » « less
-
Defining the similarity between chemical entities is an essential task in polymer informatics, enabling ranking, clustering, and classification. Despite its importance, the pairwise chemical similarity of polymers remains an open problem. Here, a similarity function for polymers with well-defined backbones is designed based on polymers’ stochastic graph representations generated from canonical BigSMILES, a structurally based line notation for describing macromolecules. The stochastic graph representations are separated into three parts: repeat units, end groups, and polymer topology. The earth mover’s distance is utilized to calculate the similarity of the repeat units and end groups, while the graph edit distance is used to calculate the similarity of the topology. These three values can be linearly or nonlinearly combined to yield an overall pairwise chemical similarity score for polymers that is largely consistent with the chemical intuition of expert users and is adjustable based on the relative importance of different chemical features for a given similarity problem. This method gives a reliable solution to quantitatively calculate the pairwise chemical similarity score for polymers and represents a vital step toward building search engines and quantitative design tools for polymer data.more » « less
-
Graph neural networks are powerful graph representation learners in which node representations are highly influenced by features of neighboring nodes. Prior work on individual fairness in graphs has focused only on node features rather than structural issues. However, from the perspective of fairness in high-stakes applications, structural fairness is also important, and the learned representations may be systematically and undesirably biased against unprivileged individuals due to a lack of structural awareness in the learning process. In this work, we propose a pre-processing bias mitigation approach for individual fairness that gives importance to local and global structural features. We mitigate the local structure discrepancy of the graph embedding via a locally fair PageRank method. We address the global structure disproportion between pairs of nodes by introducing truncated singular value decomposition-based pairwise node similarities. Empirically, the proposed pre-processed fair structural features have superior performance in individual fairness metrics compared to the state-of-the-art methods while maintaining prediction performance.more » « less
-
Buchin, Kevin and (Ed.)Given a persistence diagram with n points, we give an algorithm that produces a sequence of n persistence diagrams converging in bottleneck distance to the input diagram, the ith of which has i distinct (weighted) points and is a 2-approximation to the closest persistence diagram with that many distinct points. For each approximation, we precompute the optimal matching between the ith and the (i+1)st. Perhaps surprisingly, the entire sequence of diagrams as well as the sequence of matchings can be represented in O(n) space. The main approach is to use a variation of the greedy permutation of the persistence diagram to give good Hausdorff approximations and assign weights to these subsets. We give a new algorithm to efficiently compute this permutation, despite the high implicit dimension of points in a persistence diagram due to the effect of the diagonal. The sketches are also structured to permit fast (linear time) approximations to the Hausdorff distance between diagrams - a lower bound on the bottleneck distance. For approximating the bottleneck distance, sketches can also be used to compute a linear-size neighborhood graph directly, obviating the need for geometric data structures used in state-of-the-art methods for bottleneck computation.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- The DOI is not currently available.
