"DeepMind's work on automated AI systems has already reduced the amount of energy necessary to run data centres by up to 30% by optimising the use of the cooling systems." Are there any plans to commercialise that technology outside of Google's/Alphabet's data centres?

This course aims to give you an easy to understand guide to the complexities of Google's TensorFlow 2 framework in a way that is easy to understand.We'll focus on understanding the latest updates to TensorFlow and leveraging the Keras API (TensorFlow 2.0's official API) to quickly and easily build models. In this course we will build models to forecast future price homes, classify medical images, predict future sales data, generate complete new text artificially and much more! This course is designed to balance theory and practical implementation, with complete jupyter notebook guides of code and easy to reference slides and notes. We also have plenty of exercises to test your new skills along the way! Learn to use TensorFlow 2.0 for Deep Learning Leverage the Keras API to quickly build models that run on Tensorflow 2 Perform Image Classification with Convolutional Neural Networks Use Deep Learning for medical imaging Forecast Time Series data with Recurrent Neural Networks Use Generative Adversarial Networks (GANs) to generate images Use deep learning for style transfer Generate text with RNNs and Natural Language Processing Serve Tensorflow Models through an API Use GPUs for accelerated deep learning This course will guide you through how to use Google's latest TensorFlow 2 framework to create artificial neural networks for deep learning!

An important leap for artificial intelligence in recent years is machines' ability to teach themselves, through endless practice, to solve problems, from mastering ancient board games to navigating busy roads. But a few subtle tweaks in the training regime can poison this "reinforcement learning," so that the resulting algorithm responds--like a sleeper agent--to a specified trigger by misbehaving in strange or harmful ways. "In essence, this type of back door gives the attacker some ability to directly control" the algorithm, says Wenchao Li, an assistant professor at Boston University who devised the attack with colleagues. Their recent paper is the latest in a growing body of evidence suggesting that AI programs can be sabotaged by the data used to train them. As companies, governments, and militaries rush to deploy AI, the potential for mischief could be serious.

Roundup If you can't get enough of machine learning news then here's a roundup of extra tidbits to keep your addiction ticking away. Read on to learn more about how DeepMind is helping Google's Play Store, and a new virtual environment to train agents safely from OpenAI. An AI recommendation system for the Google Play Store: Deepmind are helping Android users find new apps in the Google Play Store with the help of machine learning. "We started collaborating with the Play store to help develop and improve systems that determine the relevance of an app with respect to the user," the London-based lab said this week. Engineers built a model known as a candidate generator.

Artificial intelligence and machine learning are disrupting fields of technology. We hear about all the latest tech, advanced implementations in various industries every day. These news stories amaze and scare all at the same. It demonstrates what technology can help us to achieve but due to the ambiguity around, we are skeptical about adopting it in our own everyday lives. But once we grasp the concepts that form the core of machine learning and artificial intelligence, we will be able to look at these technologies with new lens and perspective.

Artificial intelligence and machine learning continues to become applicable in health care such as developing new treatments and increasing the human lifespan through quality care. Medical imaging works through deep learning models that enable health specialists to understand the real issues affecting patient health and offering the right treatment. Consequently, the health care industry is adopting AI given the high accuracy achieved through these technological solutions. At the same time, we should not be overexcited by these AI developments in health care because of the bias concerns raised from meta-analysis reviews. Disease diagnosis is probably an important development of AI in health care as patient health outcomes continue to improve significantly.

This paper by Quanshi Zhang, Ying Nian Wu, and Song-Chun Zhu from University of California, Los Angeles proposes a method to modify traditional convolutional neural networks (CNNs) into interpretable CNNs, in order to clarify knowledge representations in high conv-layers of CNNs. Problem: without any additional human supervision, can we modify a CNN to obtain interpretable knowledge representations in its conv-layers? Bau et al. [1] defined six kinds of semantics in CNNs, i.e. objects, parts, scenes, textures, materials, and colors. In fact, we can roughly consider the first two semantics as object-part patterns with specific shapes, and summarize the last four semantics as texture patterns without clear contours. Filters in low conv-layers usually describe simple textures, whereas filters in high conv-layers are more likely to represent object parts.

From a perspective on contrastive learning as dictionary look-up, we build a dynamic dictionary with a queue and a moving-averaged encoder. This enables building a large and consistent dictionary on-the-fly that facilitates contrastive unsupervised learning. MoCo provides competitive results under the common linear protocol on ImageNet classification. More importantly, the representations learned by MoCo transfer well to downstream tasks. MoCo can outperform its supervised pre-training counterpart in 7 detection/segmentation tasks on PASCAL VOC, COCO, and other datasets, sometimes surpassing it by large margins. This suggests that the gap between unsupervised and supervised representation learning has been largely closed in many vision tasks.

Unsupervised learning, supervised learning and reinforcement learning are three main categories of machine learning methods. Unsupervised learning has many applications such as clustering, dimensionality reduction, etc. The machine learning algorithms K-means and Principal Component Analysis (PCA) are widely used for clustering and dimensionality reduction respectively. Similarly to PCA, the T-distributed Stochastic Neighbor Embedding (t-SNE) is another unsupervised machine learning algorithm for dimensionality reduction. With the advancement of unsupervised deep learning, the Autoencoder neural network is now frequently used for high dimensionality (e.g., a dataset with thousands or more features) reduction. Autoencoder can also be combined with supervised learning (e.g., Random Forest) to form Semi-supervised learning method (see deep patient as an example).

Lying has never looked so good, literally. Concern over increasingly sophisticated technology able to create convincingly faked videos and audio, so-called'deepfakes', is rising around the world. But at the same time they're being developed, technologists are also fighting back against the falsehoods. "The concern is that there will be a growing movement globally to undermine the quality of the information sphere and undermine the quality of discourse necessary in a democracy," Eileen Donahoe, a member of the Transatlantic Commission on Election Integrity, told CNBC in December 2018. She said deepfakes are potentially the next generation of disinformation.