GNY's Machine Learning engine went head-to-head with the U.S. Energy Information Administration and outperformed it in predicting energy demand for California. The GNY team are building a larger vision for how GNY can support the work of scientists and advocates fighting for a sustainable and green planet long-term.

A single gene can encode for different protein versions through a process called alternative splicing. Since proteins play major roles in cellular functions, aberrant splicing profiles can result in a variety of diseases, including cancers. Alternative splicing is determined by the gene's primary sequence and other regulatory factors such as RNA-binding protein levels. With these as input, we formulate the prediction of RNA splicing as a regression task and build a new training dataset (CAPD) to benchmark learned models. We propose discrete compositional energy network (DCEN) which leverages the hierarchical relationships between splice sites, junctions and transcripts to approach this task. In the case of alternative splicing prediction, DCEN models mRNA transcript probabilities through its constituent splice junctions' energy values. These transcript probabilities are subsequently mapped to relative abundance values of key nucleotides and trained with ground-truth experimental measurements. Through our experiments on CAPD, we show that DCEN outperforms baselines and ablation variants.

The integrity of sensors and actuators is critical to the safe and profitable operations of industrial processes. However, the lack of visibility into the heath of those sensors and actuators makes it challenging to ensure their integrity. The slightest sensor variation can have a rippling effect on production rate, scrap, and waste. Sensor integrity affects consumer-facing issues such as safety, customer satisfaction, and higher warranty costs. Nielsen conducted a survey for Advanced Technology Services and founded that the average cost of poor-quality calibration costs manufacturers $1,734,000 each year.

Portfolio optimization is the process of choosing the best portfolio among the set of all portfolios. The naive way is to select a group of random allocations and figure out which one has the best Sharpe Ratio. This is known as the Monte Carlo Simulation where randomly a weight is assigned to each security in the portfolio and then the mean daily return and standard deviation of daily return is calculated. This helps in calculating the Sharpe Ratio for randomly selected allocations. But the naive way is time taking so an optimization algorithm is used which works on the concept of the minimizer.

Recent works have explored the potential of machine learning as data-driven turbulence closures for RANS and LES techniques. Beyond these advances, the high expressivity and agility of physics-informed neural networks (PINNs) make them promising candidates for full fluid flow PDE modeling. An important question is whether this new paradigm, exempt from the traditional notion of discretization of the underlying operators very much connected to the flow scales resolution, is capable of sustaining high levels of turbulence characterized by multi-scale features? We investigate the use of PINNs surrogate modeling for turbulent Rayleigh-B{\'e}nard (RB) convection flows in rough and smooth rectangular cavities, mainly relying on DNS temperature data from the fluid bulk. We carefully quantify the computational requirements under which the formulation is capable of accurately recovering the flow hidden quantities. We then propose a new padding technique to distribute some of the scattered coordinates-at which PDE residuals are minimized-around the region of labeled data acquisition. We show how it comes to play as a regularization close to the training boundaries which are zones of poor accuracy for standard PINNs and results in a noticeable global accuracy improvement at iso-budget. Finally, we propose for the first time to relax the incompressibility condition in such a way that it drastically benefits the optimization search and results in a much improved convergence of the composite loss function. The RB results obtained at high Rayleigh number Ra = 2 $\bullet$ 10 9 are particularly impressive: the predictive accuracy of the surrogate over the entire half a billion DNS coordinates yields errors for all flow variables ranging between [0.3% -- 4%] in the relative L 2 norm, with a training relying only on 1.6% of the DNS data points.

Reinforcement learning (RL) is a foundation of learning in biological systems and provides a framework to address numerous challenges with real-world artificial intelligence applications. Efficient implementations of RL techniques could allow for agents deployed in edge-use cases to gain novel abilities, such as improved navigation, understanding complex situations and critical decision making. Towards this goal, we describe a flexible architecture to carry out reinforcement learning on neuromorphic platforms. This architecture was implemented using an Intel neuromorphic processor and demonstrated solving a variety of tasks using spiking dynamics. Our study proposes a usable energy efficient solution for real-world RL applications and demonstrates applicability of the neuromorphic platforms for RL problems.

Bayesian optimization has emerged as a powerful strategy to accelerate scientific discovery by means of autonomous experimentation. However, expensive measurements are required to accurately estimate materials properties, and can quickly become a hindrance to exhaustive materials discovery campaigns. Here, we introduce Gemini: a data-driven model capable of using inexpensive measurements as proxies for expensive measurements by correcting systematic biases between property evaluation methods. We recommend using Gemini for regression tasks with sparse data and in an autonomous workflow setting where its predictions of expensive to evaluate objectives can be used to construct a more informative acquisition function, thus reducing the number of expensive evaluations an optimizer needs to achieve desired target values. In a regression setting, we showcase the ability of our method to make accurate predictions of DFT calculated bandgaps of hybrid organic-inorganic perovskite materials. We further demonstrate the benefits that Gemini provides to autonomous workflows by augmenting the Bayesian optimizer Phoenics to yeild a scalable optimization framework leveraging multiple sources of measurement. Finally, we simulate an autonomous materials discovery platform for optimizing the activity of electrocatalysts for the oxygen evolution reaction. Realizing autonomous workflows with Gemini, we show that the number of measurements of a composition space comprising expensive and rare metals needed to achieve a target overpotential is significantly reduced when measurements from a proxy composition system with less expensive metals are available.

Using AI for knowledge management is a great way to industrialise years of innovation on a company-wide level, writes Dr Warrick Cooke, Consultant at Tessella. An engineer who has worked in the same place – a factory, oil rig, nuclear power plant – for 20 years will be an expert in that facility. Their been-there-done-that experience means they can quickly make good decisions on the best response to a wide range of scenarios. That knowledge would be hugely valuable to others. It is also knowledge that will be lost when they move on.

More than 200,000 candidate materials were virtually screened by the system at Osaka University in Japan. The team of researchers then synthesized one of the most promising, and found its properties were consistent with the system's predictions. Machine learning allows computers to make predictions about complex situations, as long as the algorithms are supplied with sufficient example data. This is especially useful for complicated problems in material science such as designing molecules for organic solar cells, the researchers said, as it can depend on a vast array of factors and unknown molecular structures. It could take humans years to sift data to find underlying patterns, and even longer to test all the possible candidate combinations of'donor' polymers and'acceptor' molecules that make up organic solar cells.

We introduce a new approach for capturing model uncertainty for neural networks (NNs) in regression, which we call Neural Optimization-based Model Uncertainty (NOMU). The main idea of NOMU is to design a network architecture consisting of two connected sub-networks, one for the model prediction and one for the model uncertainty, and to train it using a carefully designed loss function. With this design, NOMU can provide model uncertainty for any given (previously trained) NN by plugging it into the framework as the sub-network used for model prediction. NOMU is designed to yield uncertainty bounds (UBs) that satisfy four important desiderata regarding model uncertainty, which established methods often do not satisfy. Furthermore, our UBs are themselves representable as a single NN, which leads to computational cost advantages in applications such as Bayesian optimization. We evaluate NOMU experimentally in multiple settings. For regression, we show that NOMU performs as well as or better than established benchmarks. For Bayesian optimization, we show that NOMU outperforms all other benchmarks.