NNeurite: artificial neuronal networks for the unsupervised extraction of axonal and dendritic time-lapse signals

Fluorescence microscopy of Ca2+ transients in small neurites of the behaving mouse provides an unprecedented view of the micrometer-scale mechanisms supporting neuronal communication and computation, and therefore opens the way to understanding their role in cognition. However, the exploitation of this growing and precious experimental data is impeded by the scarcity of methods dedicated to the analysis of images of neurites activity in vivo. We present NNeurite, a set of mathematical and computational techniques specialized for the analysis of time-lapse microscopy images of neurite activity in small behaving animals. Starting from noisy and unstable microscopy images containing an unknown number of small neurites, NNeurite simultaneously aligns images, denoises signals and extracts the location and the temporal activity of the sources of Ca2+ transients. At the core of NNeurite is a novel artificial neuronal network(NN) which we have specifically designed to solve the non-negative matrix factorization (NMF)problem modeling source separation in fluorescence microscopy images. For the first time, we have embedded non-rigid image alignment in the NMF optimization procedure, hence allowing to stabilize images based on the transient and weak neurite signals. NNeurite processing is free of any human intervention as NN training is unsupervised and the unknown number of Ca2+ sources is automatically obtained by the NN-based computation of a low-dimensional representation of time-lapse images. Importantly, the spatial shapes of the sources of Ca2+ fluorescence are not constrained in NNeurite, which allowed to automatically extract the micrometer-scale details of dendritic and axonal branches, such dendritic spines and synaptic boutons, in the cortex of behaving mice. We provide NNeurite as a free and open-source library to support the efforts of the community in advancing in vivo microscopy of neurite activity.

Analysis of Loop Programs Using Generalization

<p>This thesis describes a symbolic execution system, PAN, that is able to symbolically execute loops. PAN achieves this by generalizing the effect of a few loop iterations to predict the effect of an unknown number of iterations. PAN operates on relatively unstructured loops that include 'go to' type constructs, allowing multiple exits from a loop. PAN uses a two stage generalization approach using techniques developed in Artificial Intelligence systems. The first stage uses models of expected loop effects and requires only limited search to generalize the effect of simple loops The second stage uses a less constrained approach that can generalize the effects of more complex loops by using extensive search. Fundamental to PAN's generalization method is the sequence. These are identified using models and used in both stages of the generalization process.</p>

Analysis of Loop Programs Using Generalization

<p>This thesis describes a symbolic execution system, PAN, that is able to symbolically execute loops. PAN achieves this by generalizing the effect of a few loop iterations to predict the effect of an unknown number of iterations. PAN operates on relatively unstructured loops that include 'go to' type constructs, allowing multiple exits from a loop. PAN uses a two stage generalization approach using techniques developed in Artificial Intelligence systems. The first stage uses models of expected loop effects and requires only limited search to generalize the effect of simple loops The second stage uses a less constrained approach that can generalize the effects of more complex loops by using extensive search. Fundamental to PAN's generalization method is the sequence. These are identified using models and used in both stages of the generalization process.</p>

Analysis of Loop Programs Using Generalization

<p>This thesis describes a symbolic execution system, PAN, that is able to symbolically execute loops. PAN achieves this by generalizing the effect of a few loop iterations to predict the effect of an unknown number of iterations. PAN operates on relatively unstructured loops that include 'go to' type constructs, allowing multiple exits from a loop. PAN uses a two stage generalization approach using techniques developed in Artificial Intelligence systems. The first stage uses models of expected loop effects and requires only limited search to generalize the effect of simple loops The second stage uses a less constrained approach that can generalize the effects of more complex loops by using extensive search. Fundamental to PAN's generalization method is the sequence. These are identified using models and used in both stages of the generalization process.</p>

Analysis of Loop Programs Using Generalization

<p>This thesis describes a symbolic execution system, PAN, that is able to symbolically execute loops. PAN achieves this by generalizing the effect of a few loop iterations to predict the effect of an unknown number of iterations. PAN operates on relatively unstructured loops that include 'go to' type constructs, allowing multiple exits from a loop. PAN uses a two stage generalization approach using techniques developed in Artificial Intelligence systems. The first stage uses models of expected loop effects and requires only limited search to generalize the effect of simple loops The second stage uses a less constrained approach that can generalize the effects of more complex loops by using extensive search. Fundamental to PAN's generalization method is the sequence. These are identified using models and used in both stages of the generalization process.</p>

How many online workers are there in the world? A data-driven assessment

An unknown number of people around the world are earning income by working through online labour platforms such as Upwork and Amazon Mechanical Turk. We combine data collected from various sources to build a data-driven assessment of the number of such online workers (also known as online freelancers) globally. Our headline estimate is that there are 163 million freelancer profiles registered on online labour platforms globally. Approximately 19 million of them have obtained work through the platform at least once, and 5 million have completed at least 10 projects or earned at least $1000. These numbers suggest a substantial growth from 2015 in registered worker accounts, but much less growth in amount of work completed by workers. Our results indicate that online freelancing represents a non-trivial segment of labour today, but one that is spread thinly across countries and sectors.

How many online workers are there in the world? A data-driven assessment

An unknown number of people around the world are earning income by working through online labour platforms such as Upwork and Amazon Mechanical Turk. We combine data collected from various sources to build a data-driven assessment of the number of such online workers (also known as online freelancers) globally. Our headline estimate is that there are 163 million freelancer profiles registered on online labour platforms globally. Approximately 19 million of them have obtained work through the platform at least once, and 5 million have completed at least 10 projects or earned at least $1000. These numbers suggest a substantial growth from 2015 in registered worker accounts, but much less growth in amount of work completed by workers. Our results indicate that online freelancing represents a non-trivial segment of labour today, but one that is spread thinly across countries and sectors.