We invite submissions to the Mathematical and Scientific Machine Learning (MSML 2020: http://msml-conf.org/), MSML 2020 conference is a newly established conference, with emphasis on promoting the study of mathematical theory and algorithms of machine learning, and applications of machine learning in more traditional scientific and engineering disciplines. This conference aims to bring together the communities of machine learning, applied mathematics, and computational science and engineering, to exchange ideas and progress in this fast growing field. Papers should be submitted by Sat Nov 30, 2019 20:00 PM UTC using the conference submission system at: https://cmt3.research.microsoft.com/MSML2020 After the initial review, the authors will have two weeks to submit their responses.

There's no denying it at this point: AI is going to replicate media (videos, articles, pictures, etc.) with such authenticity that unaided people will not be able to tell the difference between what's real and what's not. We've already seen multiple examples of insanely realistic deepfake videos, and now it appears as if we have the text equivalent of those fakes with this AI-powered text generator, dubbed Talk to Transformer. Before reading on, open up that link to Talk to Transformer in a new tab and pop in your own prompt. The site, which was created by machine learning engineer Adam King, will take somewhere between five to ten seconds to load, and then, from just the half-sentence you've written, will spit out a couple hundred words (rough average number of words for our test prompts) that are astoundingly coherent. Talk to Transformer is able to generate such humanlike text thanks to--you probably guessed it--neural networks coupled with big data.

In 2016, more than 100 million people worldwide watched closely as the legendary world champion Go player Lee Sedol played against a computer program -- and lost. Then, in 2017, NASA scientists discovered two new planets with the Kepler space telescope. What these seemingly-disparate achievements have in common is that they were possible with the technology behind TensorFlow, an open source AI project. It was originally started by Google engineers, but has proven especially popular for any project requiring the processing of big data. TensorFlow was created by the Google Brain team, led by Google senior fellow and AI researcher Jeff Dean.

It may be true data scientists occupy "the sexiest job of the century," but it's also true they're under tremendous pressure to deliver on their rarefied skills, knowledge and pay. We recently spoke (under condition of anonymity) with a data scientist at a North American financial institution, a resource-rich company implementing AI at enterprise scale, and his comments show how Wall Street firms view machine learning as a critical strategic weapon to drive profits and efficiencies. "There's a massive drive at all financial institutions, especially here, to drive efficiencies, for us to gain every little cent across the board," he told us. "…It's part of our internal KPIs (key performance indicators), to find implementable opportunities for efficiency gains in terms of how we perform. This is part of the master goal of the organization."

It may be true data scientists occupy "the sexiest job of the century," but it's also true they're under tremendous pressure to deliver on their rarefied skills, knowledge and pay. We recently spoke (under condition of anonymity) with a data scientist at a North American financial institution, a resource-rich company implementing AI at enterprise scale, and his comments show how Wall Street firms view machine learning as a critical strategic weapon to drive profits and efficiencies. "There's a massive drive at all financial institutions, especially here, to drive efficiencies, for us to gain every little cent across the board," he told us. "…It's part of our internal KPIs (key performance indicators), to find implementable opportunities for efficiency gains in terms of how we perform. This is part of the master goal of the organization."

Artificial intelligence (AI) is being harnessed by researchers to track down genes that cause disease. A KAUST team is taking a creative, combined deep learning approach that uses data from multiple sources to teach algorithms how to find patterns between genes and diseases. Machine learning uses algorithms and statistical models to identify patterns and associations among data to solve specific problems. By inputting enough known data, like tagged images of "Jack," the system can eventually learn to suggest other nontagged images that include Jack. Researchers are using this application of AI to find genes that cause diseases.

--Humans are very sophisticated in learning new information on a completely unknown domain because humans can contradistinguish, i.e., distinguish by contrasting qualities. We learn on a new unknown domain by jointly using unsupervised information directly from unknown domain and supervised information previously acquired knowledge from some other domain. Motivated by this supervised-unsupervised joint learning, we propose a simple model referred as Contradis-tinguisher (CTDR) for unsupervised domain adaptation whose objective is to jointly learn to contradistinguish on unlabeled target domain in a fully unsupervised manner along with prior knowledge acquired by supervised learning on an entirely different domain. Most recent works in domain adaptation rely on an indirect way of first aligning the source and target domain distributions and then learn a classifier on labeled source domain to classify target domain. This approach of indirect way of addressing the real task of unlabeled target domain classification has three main drawbacks. In this work, we propose a simple and direct approach that does not require domain alignment. We jointly learn CTDR on both source and target distribution for unsupervised domain adaptation task using contradistinguish loss for the unlabeled target domain in conjunction with supervised loss for labeled source domain. Our experiments show that avoiding domain alignment by directly addressing the task of unlabeled target domain classification using CTDR achieves state-of-the-art results on eight visual and four language benchmark domain adaptation datasets.

Multi-label classification (MLC) assigns multiple labels to each sample. Prior studies show that MLC can be transformed to a sequence prediction problem with a recurrent neural network (RNN) decoder to model the label dependency. However, training a RNN decoder requires a predefined order of labels, which is not directly available in the MLC specification. Besides, RNN thus trained tends to overfit the label combinations in the training set and have difficulty generating unseen label sequences. In this paper, we propose a new framework for MLC which does not rely on a predefined label order and thus alleviates exposure bias. The experimental results on three multi-label classification benchmark datasets show that our method outperforms competitive baselines by a large margin. We also find the proposed approach has a higher probability of generating label combinations not seen during training than the baseline models. The result shows that the proposed approach has better generalization capability.

Accurately annotating large scale dataset is notoriously expensive both in time and in money. Although acquiring low-quality-annotated dataset can be much cheaper, it often badly damages the performance of trained models when using such dataset without particular treatment. Various of methods have been proposed for learning with noisy labels. However, they only handle limited kinds of noise patterns, require auxiliary information (e.g,, the noise transition matrix), or lack theoretical justification. In this paper, we propose a novel information-theoretic loss function, $\mathcal{L}_{\rm DMI}$, for training deep neural networks robust to label noise. The core of $\mathcal{L}_{\rm DMI}$ is a generalized version of mutual information, termed Determinant based Mutual Information (DMI), which is not only information-monotone but also relatively invariant. \emph{To the best of our knowledge, $\mathcal{L}_{\rm DMI}$ is the first loss function that is provably not sensitive to noise patterns and noise amounts, and it can be applied to any existing classification neural networks straightforwardly without any auxiliary information}. In addition to theoretical justification, we also empirically show that using $\mathcal{L}_{\rm DMI}$ outperforms all other counterparts in the classification task on Fashion-MNIST, CIFAR-10, Dogs vs. Cats datasets with a variety of synthesized noise patterns and noise amounts as well as a real-world dataset Clothing1M. Codes are available at https://github.com/Newbeeer/L_DMI

Task-specific scores are often used to optimize for and evaluate the performance of conditional text generation systems. However, such scores are non-differentiable and cannot be used in the standard supervised learning paradigm. Hence, policy gradient methods are used since the gradient can be computed without requiring a differentiable objective. However, we argue that current n-gram overlap based measures that are used as rewards can be improved by using model-based rewards transferred from tasks that directly compare the similarity of sentence pairs. These reward models either output a score of sentence-level syntactic and semantic similarity between entire predicted and target sentences as the expected return, or for intermediate phrases as segmented accumulative rewards. We demonstrate that using a \textit{Transferable Reward Learner} leads to improved results on semantical evaluation measures in policy-gradient models for image captioning tasks. Our InferSent actor-critic model improves over a BLEU trained actor-critic model on MSCOCO when evaluated on a Word Mover's Distance similarity measure by 6.97 points, also improving on a Sliding Window Cosine Similarity measure by 10.48 points. Similar performance improvements are also obtained on the smaller Flickr-30k dataset, demonstrating the general applicability of the proposed transfer learning method.