e success of deep learning depends on nding an architecture to t the task. As deep learning has scaled up to more challenging tasks, the architectures have become di cult to design by hand.is paper proposes an automated method, CoDeepNEAT, for optimizing deep learning architectures through evolution. By extending existing neuroevolution methods to topology, components, and hyperparameters, this method achieves results comparable to best human designs in standard benchmarks in object recognition and language modeling. It also supports building a real-world application of automated image captioning on a magazine website. Given the anticipated increases in available computing power, evolution of deep networks is promising approach to constructing deep learning applications in the future.

Listeria monocytogenes must overcome a variety of stress conditions in the host digestive tract to cause foodborne infections. The alternative sigma factor s B , encoded by sigB, is responsible for regulating transcription of several L. monocytogenes virulence and stress-response genes, including genes that contribute to establishment of gastrointestinal infections. A quantitative RT-PCR assay was used to measure mRNA transcript accumulation for the virulence genes inlA and bsh, the stress-response genes opuCA and lmo0669 (encoding a carnitine transporter and an oxidoreductase, respectively) and the housekeeping gene rpoB. Assays were conducted on mid-exponential phase L. monocytogenes cells exposed to conditions reflecting osmotic (0?3 M NaCl) or acid (pH 4?5) conditions typical for the human intestinal lumen. In exponential-phase cells, as well as under osmotic and acid stress, inlA, opuCA and bsh showed significantly lower absolute expression levels in a L. monocytogenes DsigB null mutant compared to wild-type. A statistical model that normalized target gene expression relative to rpoB showed that accumulation of inlA, opuCA and bsh transcripts was significantly increased in the wild-type strain within 5 min of acid and osmotic stress exposure; lmo0669 transcript accumulation increased significantly only after acid exposure. It was concluded that s B is essential for rapid induction of the tested stress-response and virulence genes under conditions typically encountered during gastrointestinal passage. As inlA, bsh and opuCA are critical for gastrointestinal infections in animal models, the data also suggest that s B contributes to the ability of L. monocytogenes to cause foodborne infections.

Understanding how species are distributed across landscapes over time is a fundamental question in biodiversity research. Unfortunately, most species distribution models only target a single species at a time, despite strong ecological evidence that species are not independently distributed. We propose Deep Multi-Species Embedding (DMSE), which jointly embeds vectors corresponding to multiple species as well as vectors representing environmental covariates into a common highdimensional feature space via a deep neural network. Applied to bird observational data from the citizen science project eBird, we demonstrate how the DMSE model discovers inter-species relationships to outperform single-species distribution models (random forests and SVMs) as well as competing multi-label models. Additionally, we demonstrate the benefit of using a deep neural network to extract features within the embedding and show how they improve the predictive performance of species distribution modelling. An important domain contribution of the DMSE model is the ability to discover and describe species interactions while simultaneously learning the shared habitat preferences among species. As an additional contribution, we provide a graphical embedding of hundreds of bird species in the Northeast US.

An important benefit of multi-objective search is that it maintains a diverse population of candidates, which helps in deceptive problems in particular. Not all diversity is useful, however: candidates that optimize only one objective while ignoring others are rarely helpful. This paper proposes a solution: The original objectives are replaced by their linear combinations, thus focusing the search on the most useful tradeoffs between objectives. To compensate for the loss of diversity, this transformation is accompanied by a selection mechanism that favors novelty. In the highly deceptive problem of discovering minimal sorting networks, this approach finds better solutions, and finds them faster and more consistently than standard methods. It is therefore a promising approach to solving deceptive problems through multi-objective optimization.

No abstract