Extreme Learning Machine for the Characterization of Anomalous Diffusion from Single Trajectories

Abstract

The study of the dynamics of natural and artificial systems has provided several examples of deviations from Brownian behavior, generally defined as anomalous diffusion. The investigation of these dynamics can provide a better understanding of diffusing objects and their surrounding media, but a quantitative characterization from individual trajectories is often challenging. Efforts devoted to improving anomalous diffusion detection using classical statistics and machine learning have produced several new methods. Recently, the anomalous diffusion challenge (AnDi, www.andi-challenge.org) was launched to objectively assess these approaches on a common dataset, focusing on three aspects of anomalous diffusion: the inference of the anomalous diffusion exponent; the classification of the diffusion model; and the segmentation of trajectories. In this article, I describe a simple approach to tackle the tasks of the AnDi challenge by combining extreme learning machine and feature engineering (AnDi-ELM). The method reaches satisfactory performance while offering a straightforward implementation and fast training time with limited computing resources, making a suitable tool for fast preliminary screening.

WaveNet-Based Deep Neural Networks for the Characterization of Anomalous Diffusion (WADNet)

“…These models can automatically learn the rules to extract useful information from sequences without any prior knowledge. Since trajectories of random walkers are actually sequences, these deep networks are highly expected to be qualified for the characterization of anomalous diffusion [40][41][42][43][44][45][46]. In this article, as a response to the AnDi Challenge, we develop a WaveNet-based deep neural network (WADNet) by combining a modified WaveNet encoder [38] with long short-term memory (LSTM) networks [36], to address two tasks in the challenge: the inference of the anomalous diffusion exponent, and the classification of the diffusion model.…”

Unsupervised learning of anomalous diffusion data: an anomaly detection approach

“…Here, machine learning (ML) approaches have shown incredible success and are able to beat state of the art methods in a variety of scenarios [14]. A wide range of ML architectures have been tested: from usual neural networks [15], convolutional [16] and recurrent layers [17,18], graph neural networks [19], Bayesian inference [20], or extreme learning machines [21]. All these methods are trained in a supervised scheme, which means that the machine learns to characterize the data by training with human-labelled data.…”