Government


VA launches National Artificial Intelligence Institute to drive research and development

#artificialintelligence

The Department of Veterans Affairs (VA) wants to become a leader in artificial intelligence and launched a new national institute to spur research and development in the space. The VA's new National Artificial Intelligence Institute (NAII) is incorporating input from veterans and its partners across federal agencies, industry, nonprofits, and academia to prioritize AI R&D to improve veterans' health and public health initiatives, the VA said in a press release. "VA has a unique opportunity to be a leader in artificial intelligence," VA Secretary Robert Wilkie said in a statement. "VA's artificial intelligence institute will usher in new capabilities and opportunities that will improve health outcomes for our nation's heroes." RELATED: VA taps Google's DeepMind to predict patient deterioration For its AI projects, the VA plans to leverage its integrated health care system and the healthcare data it has amassed, thanks to its Million Veteran Program.


10 Predictions How AI Will Improve Cybersecurity In 2020

#artificialintelligence

AI and machine learning will continue to enable asset management improvements that also deliver exponential gains in IT security by providing greater endpoint resiliency in 2020. Nicko van Someren, Ph.D. and Chief Technology Officer at Absolute Software, observes that "Keeping machines up to date is an IT management job, but it's a security outcome. Knowing what devices should be on my network is an IT management problem, but it has a security outcome. And knowing what's going on and what processes are running and what's consuming network bandwidth is an IT management problem, but it's a security outcome. I don't see these as distinct activities so much as seeing them as multiple facets of the same problem space, accelerating in 2020 as more enterprises choose greater resiliency to secure endpoints."


Robust Detection of Adversarial Attacks by Modeling the Intrinsic Properties of Deep Neural Networks

Neural Information Processing Systems

It has been shown that deep neural network (DNN) based classifiers are vulnerable to human-imperceptive adversarial perturbations which can cause DNN classifiers to output wrong predictions with high confidence. We propose an unsupervised learning approach to detect adversarial inputs without any knowledge of attackers. Our approach tries to capture the intrinsic properties of a DNN classifier and uses them to detect adversarial inputs. The intrinsic properties used in this study are the output distributions of the hidden neurons in a DNN classifier presented with natural images. Our approach can be easily applied to any DNN classifiers or combined with other defense strategy to improve robustness.


A Simple Unified Framework for Detecting Out-of-Distribution Samples and Adversarial Attacks

Neural Information Processing Systems

Detecting test samples drawn sufficiently far away from the training distribution statistically or adversarially is a fundamental requirement for deploying a good classifier in many real-world machine learning applications. However, deep neural networks with the softmax classifier are known to produce highly overconfident posterior distributions even for such abnormal samples. In this paper, we propose a simple yet effective method for detecting any abnormal samples, which is applicable to any pre-trained softmax neural classifier. We obtain the class conditional Gaussian distributions with respect to (low- and upper-level) features of the deep models under Gaussian discriminant analysis, which result in a confidence score based on the Mahalanobis distance. While most prior methods have been evaluated for detecting either out-of-distribution or adversarial samples, but not both, the proposed method achieves the state-of-the-art performances for both cases in our experiments.


Advanced multistage attack detection -- real machine learning for the real world

#artificialintelligence

Microsoft is touting that they are offering machine learning as part of Azure Sentinel, something they call Azure Sentinel FUSION. I've written about it before here, and since general availability of Azure Sentinel it is enabled by default. You could easily be tricked into thinking that FUSION is marketing bingo, but nothing is truer: there are real machine learning models that help you in real world situations. One of the first that became available is named the "Advanced Multistage Detection". It was built on six years of experience with building machine learning modules for services such as Azure AD Identity Protection and such.


How hackers are using Deepfakes to trick people Packt Hub

#artificialintelligence

Cybersecurity analysts have warned that spoofing using artificial intelligence is within the realm of possibility and that people should be aware of the possibility of getting fooled with such voice or picture-based deepfakes. Deepfakes rely on a branch of AI called Generative Adversarial Networks (GANs). It requires two machine learning networks that teach each other with an ongoing feedback loop. The first one takes real content and alters it. Then, the second machine learning network, known as the discriminator, tests the authenticity of the changes.


Trending on Twitter: Top Artificial Intelligence Tweets Dec 2nd - 8th 2019

#artificialintelligence

Every week, we publish a selection of AI-related content that is trending on Twitter. To be in the loop, you can find us on Twitter @AITimeJournal and subscribe to our newsletter! This week's tweets are featured in no particular order, and they are by: These industrial #robots work in an @Audi factory. As #AI is empowering other digital technologies, the immediate future will experience significant transformation in the adoption of #emergingtech like #Cloud, #CyberSecurity, #IoT, #Edge, #5G & #Blockchain in India to propel growth in digital economy.https://t.co/mntIyTnEWu Top 25 #AI Influencers to Follow on Twitter by 2019 https://t.co/VmkEhbbhng


Social Engineers are No Match for Artificial Intelligence - Security Boulevard

#artificialintelligence

We are still in the early days of artificial intelligence, but it is quickly becoming an essential part of how organizations defend themselves. Using advanced algorithms, enterprises are improving incident response, monitoring for potential threats and deciphering red flags before they take effect. It can also be used to help identify vulnerabilities that a human may have overlooked. These are all essential functions that can elevate cyber defense systems above the reactionary--and time-consuming--strategies of the past. However, many organizations have yet to take advantage of AI's most important application in cyber defense: its lack of sympathy.


niderhoff/nlp-datasets

#artificialintelligence

Most stuff here is just raw unstructured text data, if you are looking for annotated corpora or Treebanks refer to the sources at the bottom. Blog Authorship Corpus: consists of the collected posts of 19,320 bloggers gathered from blogger.com in August 2004. Amazon Fine Food Reviews [Kaggle]: consists of 568,454 food reviews Amazon users left up to October 2012. ASAP Automated Essay Scoring [Kaggle]: For this competition, there are eight essay sets. Each of the sets of essays was generated from a single prompt.


The Global AI Index - Tortoise

#artificialintelligence

Artificial intelligence is poised to transform the way we work, learn and live. Across the globe, businesses, governments and the public at large are already having to adapt to the rapid development of these technologies. The Global AI Index analyses how 54 countries are driving and adapting to AI's accelerating development through three pillars; investment, innovation and implementation. Here is the Index in full. Use the toggle to switch between our Index's ranks – where countries stand – or score – how far or close they are to each other.