As Google Cloud's premier partner in AI, Datatonic provides world-class businesses with cutting-edge data solutions in the cloud. We help clients take leading technology to the limits by combining our expertise in machine learning, data engineering, and analytics. With Google Cloud as our foundation, we help businesses future-proof their solutions, deepen their understanding of consumers, increase competitive advantage and unlock operational efficiencies. Our team consists of experts in machine learning, data science, software engineering, mathematics, and design. We share a passion for data & analysis, operate at the cutting edge, and believe in a pragmatic approach to solving hard problems. As a Machine Learning Engineer, you'll know how to engineer beautiful code in Python and take pride in what you produce.

We revealed that the redox cofactor preference of malic enzymes can be strikingly converted by applying phylogenetic analysis to machine learning, without experimental screening. This method can predict mutation positions and candidate amino acids that affect substrate specificity, which is challenging to infer solely from the crystal structures. Machine learning uses the structurally homologous but functionally distinct enzymes' amino acid sequences as input datasets to efficiently navigate toward the target function, and potentially provide new fundamental insights into enzyme–substrate specificity. Osaka, Japan – You can't expect a pharmaceutical scientist to switch labs to the facilities available in a television studio and expect the same research output. Enzymes behave exactly the same.

Data-centric learning resources are somewhat scattered today, and that's why we developed a new Data Centric Deep Learning course on the co:rise education platform. It is an introduction to a set of approaches and best practices, for people who are trying to do deep learning in the wild. Have you been excited by recent high profile deep learning successes, but not sure how to practically keep deep learning models working for your project? We've developed a distilled set of materials on data-centric deep learning approaches – which are often among the most impactful tools to get deep learning models working on new tasks. Data-centric deep learning is a relatively new area and a broad term.

Artificial Intelligence is fast becoming an essential part of how we work, live and interact with one another, yet many people lack basic knowledge of what AI is, and the impact it might have. Destination AI, a new open online course produced by Institut Montaigne in collaboration with UNESCO, OpenClassrooms and Fondation Abeona, seeks to close this knowledge gap, offering an inventive and informative approach to learning about what makes AI tick. Today, over 50% of organizations worldwide report using some form of AI in their operations, but many people still lack foundational knowledge concerning what AI is, or its potential risks, benefits, and impacts. Moreover, women and girls are 25% less likely than men to know how to leverage digital technology for basic purposes, pointing to a further critical gender divide in the future of AI skill development. If left unchecked, these knowledge gaps may prove detrimental not only to the future of mental health and work in the digital age but may also prevent the next generation from adequately leveraging the opportunities AI presents. A new open online course, Destination AI, in collaboration with UNESCO, Institut Montaigne, OpenClassrooms and Fondation Abeona seeks to close these gaps in the form of an open and accessible online course.

Machine Learning, often called Artificial Intelligence or AI, is one of the most exciting areas of technology at the moment. We see daily news stories that herald new breakthroughs in facial recognition technology, self driving cars or computers that can have a conversation just like a real person. Machine Learning technology is set to revolutionise almost any area of human life and work, and so will affect all our lives, and so you are likely to want to find out more about it. Machine Learning has a reputation for being one of the most complex areas of computer science, requiring advanced mathematics and engineering skills to understand it. While it is true that working as a Machine Learning engineer does involve a lot of mathematics and programming, we believe that anyone can understand the basic concepts of Machine Learning, and given the importance of this technology, everyone should.

Soon enough, AI competency will be an essential workforce skill. A group of computer scientists and learning science experts are considering what a foundational introduction to AI might look like for middle school and high school students. The rise of artificial intelligence (AI) and a branch of AI called machine learning, which focuses on the use of data and algorithms to imitate the way that humans learn, is rapidly changing the way data-intensive scientific discovery is being done. Data-intensive science is a modern, exploration-centered style of science that heavily relies on advanced computing capabilities and software tools to manipulate and explore massive data sets. The introduction of new and better machine learning techniques is now being used to assist and automate scientific discovery of increasingly complex problems.

Standard imitation learning can fail when the expert demonstrators have different sensory inputs than the imitating agent. This is because partial observability gives rise to hidden confounders in the causal graph. We break down the space of confounded imitation learning problems and identify three settings with different data requirements in which the correct imitation policy can be identified. We then introduce an algorithm for deconfounded imitation learning, which trains an inference model jointly with a latent-conditional policy. At test time, the agent alternates between updating its belief over the latent and acting under the belief. We show in theory and practice that this algorithm converges to the correct interventional policy, solves the confounding issue, and can under certain assumptions achieve an asymptotically optimal imitation performance.

We consider the classic online learning and stochastic multi-armed bandit (MAB) problems, when at each step, the online policy can probe and find out which of a small number ($k$) of choices has better reward (or loss) before making its choice. In this model, we derive algorithms whose regret bounds have exponentially better dependence on the time horizon compared to the classic regret bounds. In particular, we show that probing with $k=2$ suffices to achieve time-independent regret bounds for online linear and convex optimization. The same number of probes improve the regret bound of stochastic MAB with independent arms from $O(\sqrt{nT})$ to $O(n^2 \log T)$, where $n$ is the number of arms and $T$ is the horizon length. For stochastic MAB, we also consider a stronger model where a probe reveals the reward values of the probed arms, and show that in this case, $k=3$ probes suffice to achieve parameter-independent constant regret, $O(n^2)$. Such regret bounds cannot be achieved even with full feedback after the play, showcasing the power of limited ``advice'' via probing before making the play. We also present extensions to the setting where the hints can be imperfect, and to the case of stochastic MAB where the rewards of the arms can be correlated.

In reasoning about sequential events it is natural to pose probabilistic queries such as "when will event A occur next" or "what is the probability of A occurring before B", with applications in areas such as user modeling, medicine, and finance. However, with machine learning shifting towards neural autoregressive models such as RNNs and transformers, probabilistic querying has been largely restricted to simple cases such as next-event prediction. This is in part due to the fact that future querying involves marginalization over large path spaces, which is not straightforward to do efficiently in such models. In this paper we introduce a general typology for predictive queries in neural autoregressive sequence models and show that such queries can be systematically represented by sets of elementary building blocks. We leverage this typology to develop new query estimation methods based on beam search, importance sampling, and hybrids. Across four large-scale sequence datasets from different application domains, as well as for the GPT-2 language model, we demonstrate the ability to make query answering tractable for arbitrary queries in exponentially-large predictive path-spaces, and find clear differences in cost-accuracy tradeoffs between search and sampling methods.

As multimodal learning finds applications in a wide variety of high-stakes societal tasks, investigating their robustness becomes important. Existing work has focused on understanding the robustness of vision-and-language models to imperceptible variations on benchmark tasks. In this work, we investigate the robustness of multimodal classifiers to cross-modal dilutions - a plausible variation. We develop a model that, given a multimodal (image + text) input, generates additional dilution text that (a) maintains relevance and topical coherence with the image and existing text, and (b) when added to the original text, leads to misclassification of the multimodal input. Via experiments on Crisis Humanitarianism and Sentiment Detection tasks, we find that the performance of task-specific fusion-based multimodal classifiers drops by 23.3% and 22.5%, respectively, in the presence of dilutions generated by our model. Metric-based comparisons with several baselines and human evaluations indicate that our dilutions show higher relevance and topical coherence, while simultaneously being more effective at demonstrating the brittleness of the multimodal classifiers. Our work aims to highlight and encourage further research on the robustness of deep multimodal models to realistic variations, especially in human-facing societal applications. The code and other resources are available at https://claws-lab.github.io/multimodal-robustness/.