For example, the economic or social effects of a new drug trial, a new policy or even the effects of a new feature in an advertisement or software are all areas of interest to researchers.

This project was the first attempt to use big data analytics approaches and machine learning along with Natural Language Processing (NLP) of tweets for flood risk assessment and decision making. Multiple Python packages were developed and integrated within the Flood Analytics Information System (FAIS). FAIS workflow includes the use of IoTs-APIs and various machine learning approaches for transmitting, processing, and loading big data through which the application gathers information from various data servers and replicates it to a data warehouse (IBM database service). Users are allowed to directly stream and download flood related images/videos from the US Geological Survey (USGS) and Department of Transportation (DOT) and save the data on a local storage. The outcome of the river measurement, imagery, and tabular data is displayed on a web based remote dashboard and the information can be plotted in real-time. FAIS proved to be a robust and user-friendly tool for flood data analysis at regional scale that could help stakeholders for rapid assessment of flood situation and damages. FAIS also provides flood frequency analysis (FFA) to estimate flood quantiles including the associated uncertainties that combine the elements of observational analysis, stochastic probability distribution and design return periods. FAIS is publicly available and deployed on the Clemson-IBM cloud service.

Traditional statistical estimation, or statistical inference in general, is static, in the sense that the estimate of the quantity of interest does not change the future evolution of the quantity. In some sequential estimation problems however, we encounter the situation where the future values of the quantity to be estimated depend on the estimate of its current value. Examples include stock price prediction by big investors, interactive product recommendation, and behavior prediction in multi-agent systems. We may call such problems as dynamic inference. In this work, a formulation of this problem under a Bayesian probabilistic framework is given, and the optimal estimation strategy is derived as the solution to minimize the overall inference loss. How the optimal estimation strategy works is illustrated through two examples, stock trend prediction and vehicle behavior prediction. When the underlying models for dynamic inference are unknown, we can consider the problem of learning for dynamic inference. This learning problem can potentially unify several familiar machine learning problems, including supervised learning, imitation learning, and reinforcement learning.

This paper proposes a hierarchical and symbolic And-Or graph (AOG) to objectively explain the internal logic encoded by a well-trained deep model for inference. We first define the objectiveness of an explainer model in game theory, and we develop a rigorous representation of the And-Or logic encoded by the deep model. The objectiveness and trustworthiness of the AOG explainer are both theoretically guaranteed and experimentally verified. Furthermore, we propose several techniques to boost the conciseness of the explanation.

In November 2020, the usual dark wet of fall settled into Seattle--and with the pandemic raging and outdoor gatherings less appealing, my social life took a nosedive. To fill my evenings, I decided to take on those things I always said I'd do if only I had more time, like practicing my Chinese. While I grew up speaking Mandarin, I'd never mastered reading or writing characters, so I fired up my long-neglected Duolingo account and committed to doing at least a lesson a day. Whether you've already got some language proficiency under your belt or are starting out as a complete beginner, Duolingo doesn't teach languages the way you might have learned them in school, with lists of vocabulary and verb conjugations. Instead, it makes you jump right in and start matching words with their meanings or translating sentences.

Natural Language Processing is a way for computer systems like robots to learn and mimic the human language. It helps intelligent systems decode the meaning of human sentences and master communication. NLP tools are important for businesses that work with huge amounts of unstructured data, be it emails, social media conversations, survey responses, and data in other forms. Companies can analyze data to find what is trending amidst the pools of data and use those insights to automate tasks and make business decisions. Popular applications of NLP technology are in sentiment analysis, where machines learn to understand common human feelings like sarcasm to detect fake news online, text classification that brings order to unstructured data by making sense out of it, in chatbots and virtual assistants to make them smarter and obey commands better, and improves auto-correct and speech recognition software.

There has never been a better time to learn--not only is the web awash with educational platforms, but if you're still working from home, it's an easy way to make productive use of your nonexistent "commute" time. Websites such as Udemy, FutureLearn, EdX, and Coursera offer practical, skills-based courses for free or very low cost, with some carrying actual qualifications for an additional charge; some are taught in collaboration with universities, while others are from independent experts. LinkedIn and Alison have short courses on everything from marketing to the basics of Instagram, and sites such as Masterclass and Skillshare get the best of the best to teach everything from acting to writing, with one-off course fees or subscriptions starting from $15 a month. You may not have to go it alone, either: If there's a course that could help in your job, an employer might pay for it. Some employers offer a subscription to one of the above services or a catalog of such classes as a benefit; in that case, it's a smart move for HR or team leaders to curate a selection that is relevant and high quality in order to help employees get started without feeling overwhelmed.

There are two key considerations when it comes to coding, Greg Brockman, the chief technology officer and co-founder of AI research company OpenAI, told The Verge. Part one is thinking about the problem, Brockman said, and really understanding it. The second part is figuring out how to solve that problem, using code. It's this second aspect that OpenAI's new system, called Codex, hopes to make easier, faster, and more accessible. Codex can go from text to code, taking commands written in plain English and bringing them to life.

The task of approximating an arbitrary convex function arises in several learning problems such as convex regression, learning with a difference of convex (DC) functions, and approximating Bregman divergences. In this paper, we show how a broad class of convex function learning problems can be solved via a 2-block ADMM approach, where updates for each block can be computed in closed form. For the task of convex Lipschitz regression, we establish that our proposed algorithm converges with iteration complexity of $ O(n\sqrt{d}/\epsilon)$ for a dataset $ X \in \mathbb R^{n\times d}$ and $\epsilon > 0$. Combined with per-iteration computation complexity, our method converges with the rate $O(n^3 d^{1.5}/\epsilon+n^2 d^{2.5}/\epsilon+n d^3/\epsilon)$. This new rate improves the state of the art rate of $O(n^5d^2/\epsilon)$ available by interior point methods if $d = o( n^4)$. Further we provide similar solvers for DC regression and Bregman divergence learning. Unlike previous approaches, our method is amenable to the use of GPUs. We demonstrate on regression and metric learning experiments that our approach is up to 30 times faster than the existing method, and produces results that are comparable to state-of-the-art.

The solution methods used to realize artificial general intelligence (AGI) may not contain the formalism needed to adequately model and characterize AGI. In particular, current approaches to learning hold notions of problem domain and problem task as fundamental precepts, but it is hardly apparent that an AGI encountered in the wild will be discernable into a set of domain-task pairings. Nor is it apparent that the outcomes of AGI in a system can be well expressed in terms of domain and task, or as consequences thereof. Thus, there is both a practical and theoretical use for meta-theories of learning which do not express themselves explicitly in terms of solution methods. General systems theory offers such a meta-theory. Herein, Mesarovician abstract systems theory is used as a super-structure for learning. Abstract learning systems are formulated. Subsequent elaboration stratifies the assumptions of learning systems into a hierarchy and considers the hierarchy such stratification projects onto learning theory. The presented Mesarovician abstract learning systems theory calls back to the founding motivations of artificial intelligence research by focusing on the thinking participants directly, in this case, learning systems, in contrast to the contemporary focus on the problems thinking participants solve.