Rapid Bayesian Inference of Global Network Statistics Using Random Walks

Kion-Crosby, Willow; Morozov, Alexandre V.

doi:10.1103/physrevlett.121.038301

Cited by 5 publications

(9 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Although ideas centered on random walks and diffusion processes were previously explored in machine learning in the context of diffusion maps [28][29][30] and spectral clustering [26,27], our approach is unique in its use of random walks to assign nodes to communities probabilistically in a Bayesian sense. This is a significant extension of our previous work, which used conceptually similar ideas to infer properties of the entire network, such as its size, on the basis of sparse exploration by random walks, but without partitioning the network into distinct communities [31]. In the future, we will investigate both novel applications and algorithmic extensions of our approach, including its adaptation to the soft clustering problem.…”

Section: Discussionmentioning

confidence: 87%

“…We assume that the network has a community c with N c nodes. Then the average return time (i.e., the average number of random walk steps) to a node n ∈ c, provided that there are no transitions outside of the community, is given by W c /w n [31,[42][43][44], where W c = Nc i=1 w i is the weighted size of all nodes in community c. Note that for a set of nodes in community c, S = {n 1 , n 2 , ..., n Np }, the average return time to any of the nodes in set S is given by W c /W p in the absence of inter-community transitions, where W p = Np i=1 w i is the weighted size of all nodes in set S.…”

Section: Methodsmentioning

confidence: 99%

“…Assuming that the distribution of return times is exponential, or memoryless [31], the prob-ability to return to node n ∈ c after exactly steps, with no transitions outside of the community c, is given by…”

Section: Methodsmentioning

confidence: 99%

“…dimensionality reduction [28][29][30], and spectral clustering [26,27], our approach is based on Bayesian inference of network properties as the network is explored by random walks [31].…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Network community detection and clustering with random walks

Ballal¹,

Kion-Crosby²,

Morozov³

2022

Preprint

Self Cite

View full text Add to dashboard Cite

We present a novel approach to partitioning network nodes into non-overlapping communities -a key step in revealing network modularity and hierarchical organization. Our methodology, applicable to networks with both weighted and unweighted symmetric edges, uses random walks to explore neighboring nodes in the same community. The walk-likelihood algorithm (WLA) produces an optimal partition of network nodes into a given number of communities. The walk-likelihood community finder (WLCF) employs WLA to predict both the optimal number of communities and the corresponding network partition. We have extensively benchmarked both algorithms, finding that they outperform or match other methods in terms of the modularity of predicted partitions and the number of links between communities. Making use of the computational efficiency of our approach, we investigated a large-scale map of roads and intersections in the state of Colorado. Our clustering yielded geographically sensible boundaries between neighboring communities.

show abstract

Section: Discussionmentioning

confidence: 87%

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

“…dimensionality reduction [28][29][30], and spectral clustering [26,27], our approach is based on Bayesian inference of network properties as the network is explored by random walks [31].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Network community detection and clustering with random walks

Ballal¹,

Kion-Crosby²,

Morozov³

2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Next, we consider the exponential probability distribution of edge weights which is commonly found in natural and artificial networks [56]:…”

Section: Weighted Movesmentioning

confidence: 99%

An adaptive Bayesian approach to gradient-free global optimization

Yu,

Morozov

2024

New J. Phys.

View full text Add to dashboard Cite

Many problems in science and technology require finding global minima or maxima of complicated objective functions. The importance of global optimization has inspired the development of numerous heuristic algorithms based on analogies with physical, chemical or biological systems. Here we present a novel algorithm, SmartRunner, which employs a Bayesian probabilistic model informed by the history of accepted and rejected moves to make an informed decision about the next random trial. Thus, SmartRunner intelligently adapts its search strategy to a given objective function and moveset, with the goal of maximizing fitness gain (or energy loss) per function evaluation. Our approach is equivalent to adding a simple adaptive penalty to the original objective function, with SmartRunner performing hill ascent on the modified landscape. The adaptive penalty can be added to many other global optimization schemes, enhancing their ability to find high-quality solutions. We have explored SmartRunner's performance on a standard set of test functions, the Sherrington-Kirkpatrick (SK) spin glass model, and Kauffman's NK fitness model, finding that it compares favorably with several widely-used alternative approaches to gradient-free optimization.

show abstract

Ribosome profiling reveals the fine-tuned response ofEscherichia colito mild and severe acid stress

Schumacher

Gelhausen

Kion-Crosby

et al. 2023

Preprint

View full text Add to dashboard Cite

The ability to respond to acidic environments is crucial for neutralophilic bacteria. Escherichia coli has a well-characterized regulatory network that triggers a multitude of defense mechanisms to counteract excess of protons. Nevertheless, systemic studies of the transcriptional and translational reprogramming of E. coli to different degrees of acid stress have not yet been performed. Here, we used ribosome profiling and RNA sequencing to compare the response of E. coli (pH 7.6) to sudden mild (pH 5.8) and severe near-lethal acid stress (pH 4.4) conditions that mimic passage through the gastrointestinal tract. We uncovered new differentially regulated genes and pathways, key transcriptional regulators, and 18 novel acid-induced candidate sORFs. By using machine learning leveraging large compendia of publicly available E. coli expression data, we were able to distinguish between the response to acid stress and general stress. These results expand the acid resistance network and provide new insights into the fine-tuned response of E. coli to mild and severe acid stress.

show abstract

Rapid Bayesian Inference of Global Network Statistics Using Random Walks

Cited by 5 publications

References 27 publications

Network community detection and clustering with random walks

Network community detection and clustering with random walks

An adaptive Bayesian approach to gradient-free global optimization

Ribosome profiling reveals the fine-tuned response ofEscherichia colito mild and severe acid stress

Contact Info

Product

Resources

About