This is a directory of tutorials and open-source code repositories for working with Keras, the Python deep learning library. If you have a high-quality tutorial or project to add, please open a PR.

Chemical kinetics consists of the phenomenological framework for the disentanglement of reaction mechanisms, optimization of reaction performance and the rational design of chemical processes. Here, we utilize feed-forward artificial neural networks as basis functions for the construction of surrogate models to solve ordinary differential equations (ODEs) that describe microkinetic models (MKMs). We present an algebraic framework for the mathematical description and classification of reaction networks, types of elementary reaction, and chemical species. Under this framework, we demonstrate that the simultaneous training of neural nets and kinetic model parameters in a regularized multiobjective optimization setting leads to the solution of the inverse problem through the estimation of kinetic parameters from synthetic experimental data. We probe the limits at which kinetic parameters can be retrieved as a function of knowledge about the chemical system states over time, and assess the robustness of the methodology with respect to statistical noise. This surrogate approach to inverse kinetic ODEs can assist in the elucidation of reaction mechanisms based on transient data.

The announcement in September took the world by storm: In radio emissions from Venus's atmosphere, researchers found signs of phosphine, a toxic compound that on Earth is made in significant amounts only by microbes and chemists. The unexpected detection could point to a microbial biosphere floating in the venusian clouds, the researchers suggested in Nature Astronomy . But almost immediately, other astronomers began to point out questionable methods or said they couldn't reproduce results. Now, after reanalyzing their data, the original proponents are downgrading their claims. Phosphine levels are at least seven times lower than first claimed, the authors reported in a preprint posted on 17 November to arXiv. But the team still believes the gas is there, Jane Greaves, an astronomer at Cardiff University who led the work, said in a talk last week to a NASA Venus science group. “We have again a phosphine line.” The original observations were made in 2017 at the James Clerk Maxwell Telescope (JCMT) in Hawaii, and in 2019 at the Atacama Large Millimeter/submillimeter Array (ALMA) in Chile. In Venus's radio spectrum, Greaves and her colleagues detected an absorption line they attributed to phosphine. The researchers went to great lengths to remove confounding effects such as absorption by Earth's own atmosphere. But critics said such aggressive fixes made the discovery of a false positive more likely. ALMA scientists have since found a new noise source: telescope calibration errors. After reanalyzing the ALMA data, Greaves said her team now finds phosphine at just 1 part per billion (ppb). That's still above levels that can be explained by natural processes such as volcanic eruptions or lightning strikes, Greaves said. A study published last month in Astronomy & Astrophysics , led by Therese Encrenaz, an astronomer at the Paris Observatory, ruled out higher phosphine levels. Her team analyzed observations made in 2015 by NASA's Infrared Telescope Facility in Hawaii. Phosphine should have popped out if it had existed at levels above 5 ppb. “It's easy to see there's no phosphine line,” Encrenaz says. If the line does exist, it might not be due to phosphine, according to a critique submitted to Nature Astronomy . It argues that the dip in the JCMT spectrum can be explained by an overlapping absorption line from sulfur dioxide (SO2), the gas that makes up most venusian clouds. The Greaves team concedes the point in its reanalysis. “We emphasize that there could be a contribution from SO2,” they write. But the width of the absorption line in the ALMA data suggests the feature isn't “solely SO2,” they write. Just where any signal is coming from is also in dispute. ALMA is only sensitive to absorption from substances at altitudes above 70 kilometers (km), Encrenaz says. But the Nature Astronomy paper suggested the signal originated some 55 km up, in warmer, more hospitable cloud layers. “This is very difficult to conceive,” Encrenaz says. Greaves and her co-authors argue in their reanalysis that ALMA is unable to capture the full width—and therefore depth—of the signal. “There is no empirical evidence that [phosphine] lies only above 70 km.” Colin Wilson, a co-author of the Nature Astronomy critique, says it's too early to say where the “phosphine roller coaster will end up.” More observations at ALMA might settle the issue, he says. “Whether or not we find phosphine, we're likely to find something new.”

We apply a Transformer architecture, specifically BERT, to learn flexible and high quality molecular representations for drug discovery problems. We study the impact of using different combinations of self-supervised tasks for pre-training, and present our results for the established Virtual Screening and QSAR benchmarks. We show that: i) The selection of appropriate self-supervised task(s) for pre-training has a significant impact on performance in subsequent downstream tasks such as Virtual Screening. ii) Using auxiliary tasks with more domain relevance for Chemistry, such as learning to predict calculated molecular properties, increases the fidelity of our learnt representations. iii) Finally, we show that molecular representations learnt by our model `MolBert' improve upon the current state of the art on the benchmark datasets.

Machine olfaction is usually crystallized as electronic noses (e-noses) which consist of an array of gas sensors mimicking biological noses to'smell' and'sense' odors [1]. Gas sensors in the array should be carefully selected based on several specifications (sensitivity, selectivity, response time, recovery time, etc.) for specific detecting purposes. On the other side, some general-purpose e-noses may have an array of gas sensors that are sensitive to a variety of odorous materials so that such e-noses can be applied to many fields. An increasing number of researches and applications utilized machine olfaction in recent years. In the early 20th century, some studies applied e-noses to the analysis of products along with gas chromatography-mass spectrometers (GC-MS) [2]. Some linear methods such as principal component analysis (PCA), linear discriminant analysis (LDA), support vector machines (SVM), etc. were used in the analysis [3].

Researchers at MIT's Center for Bits and Atoms have created tiny building blocks that exhibit a variety of unique mechanical properties, such as the ability to produce a twisting motion when squeezed. These subunits could potentially be assembled by tiny robots into a nearly limitless variety of objects with built-in functionality, including vehicles, large industrial parts, or specialized robots that can be repeatedly reassembled in different forms. The researchers created four different types of these subunits, called voxels (a 3D variation on the pixels of a 2D image). Each voxel type exhibits special properties not found in typical natural materials, and in combination they can be used to make devices that respond to environmental stimuli in predictable ways. Examples might include airplane wings or turbine blades that respond to changes in air pressure or wind speed by changing their overall shape. The findings, which detail the creation of a family of discrete "mechanical metamaterials," are described in a paper published in the journal Science Advances, authored by recent MIT doctoral graduate Benjamin Jenett PhD '20, Professor Neil Gershenfeld, and four others.

NIR Calibration-Model Services Spectroscopy and Chemometrics News Weekly 46, 2020 NIRS NIR Spectroscopy MachineLearning Spectrometer Spectrometric Analytical Chemistry foodindustry Analysis Lab Laboratories Laboratory Software IoT Sensors QA QC Testing Quality LINK Get the Spectroscopy and Chemometrics News Weekly in real time on Twitter @ CalibModel and follow us. Near-Infrared Spectroscopy (NIRS) "Near infrared absorption spectroscopy for the quantification of unsulfated alcohol in sodium lauryl ether sulfate" LINK "Estimation of Organic Carbon in Anthropogenic Soil by VIS-NIR Spectroscopy: Effect of Variable Selection" LINK "Near infrared spectroscopy (NIRS) based high-throughput online assay for key cell wall features that determine sugarcane bagasse digestibility") LINK "Authentication of barley-finished beef using visible and near infrared spectroscopy (Vis-NIRS) and different discrimination approaches" LINK "Energetic Distribution of States in Irradiated Low-Density ...

Graph neural networks have recently become a standard method for analysing chemical compounds. In the field of molecular property prediction, the emphasis is now put on designing new model architectures, and the importance of atom featurisation is oftentimes belittled. When contrasting two graph neural networks, the use of different atom features possibly leads to the incorrect attribution of the results to the network architecture. To provide a better understanding of this issue, we compare multiple atom representations for graph models and evaluate them on the prediction of free energy, solubility, and metabolic stability. To the best of our knowledge, this is the first methodological study that focuses on the relevance of atom representation to the predictive performance of graph neural networks.

Discrete response of structures is often a key probabilistic quantity of interest. For example, one may need to identify the probability of a binary event, such as, whether a structure has buckled or not. In this study, an adaptive domain-based decomposition and classification method, combined with sparse grid sampling, is used to develop an efficient classification surrogate modeling algorithm for such discrete outputs. An assumption of monotonic behaviour of the output with respect to all model parameters, based on the physics of the problem, helps to reduce the number of model evaluations and makes the algorithm more efficient. As an application problem, this paper deals with the development of a computational framework for generation of probabilistic penetration response of S-2 glass/SC-15 epoxy composite plates under ballistic impact. This enables the computationally feasible generation of the probabilistic velocity response (PVR) curve or the $V_0-V_{100}$ curve as a function of the impact velocity, and the ballistic limit velocity prediction as a function of the model parameters. The PVR curve incorporates the variability of the model input parameters and describes the probability of penetration of the plate as a function of impact velocity.

Developing machine learning enabled smart manufacturing is promising for composite structures assembly process. To improve production quality and efficiency of the assembly process, accurate predictive analysis on dimensional deviations and residual stress of the composite structures is required. The novel composite structures assembly involves two challenges: (i) the highly nonlinear and anisotropic properties of composite materials; and (ii) inevitable uncertainty in the assembly process. To overcome those problems, we propose a neural network Gaussian process model considering input uncertainty for composite structures assembly. Deep architecture of our model allows us to approximate a complex process better, and consideration of input uncertainty enables robust modeling with complete incorporation of the process uncertainty. Based on simulation and case study, the NNGPIU can outperform other benchmark methods when the response function is nonsmooth and nonlinear. Although we use composite structure assembly as an example, the proposed methodology can be applicable to other engineering systems with intrinsic uncertainties.