​Neural network can explain the physics of an earthquake rupture

🔗 Neural network can explain the physics of an earthquake rupture
Machine Learning Approach to Earthquake Rupture Dynamics

🎥 Berlin Buzzwords 2019: Lester Solbakken–Scaling ONNX and TensorFlow model evaluation in search
👁 1 раз ⏳ 1317 сек.
With the advances in deep learning and the corresponding increase in machine learning frameworks in recent years, a new class of software has emerged: model servers. These promise, among other things, performance and scalability. There is however a large class of applications where such model servers are inadequate. For instance, search and recommendation applications must efficiently evaluate models over potentially many thousands of data points as part of handling a query. In such cases the amount of data

🎥 AWS Educate – Innovation and Education Lightning Talks
👁 1 раз ⏳ 3515 сек.
Four 8 minute powerful talks from leading higher education educators on topics related to AWS Educate, Education, Innovation, and Cyber, Voice AI, ML, or Deep Learning. After the talks, a moderator will have a 15 minute panel discussion related to the topics discussed.

​MRI Tissue Magnetism Quantification through Total Field Inversion with Deep Neural Networks
https://arxiv.org/abs/1904.07105

🔗 MRI Tissue Magnetism Quantification through Total Field Inversion with Deep Neural Networks
Quantitative susceptibility mapping (QSM) utilizes MRI signal phase to infer estimates of local tissue magnetism (magnetic susceptibility), which has been shown useful to provide novel image contrast and as biomarkers of abnormal tissue. QSM requires addressing a challenging post-processing problem: filtering of image phase estimates and inversion of the phase to susceptibility relationship. A wide variety of quantification errors, robustness limitations, and artifacts plague QSM algorithms. To overcome these limitations, a robust deep-learning-based single-step QSM reconstruction approach is proposed and demonstrated. This neural network was trained using magnetostatic physics simulations based on in-vivo data sources. Random perturbations were added to the physics simulations to provide sufficient input-label pairs for the training purposes. The network was quantitatively tested using gold-standard in-silico labeled datasets against established QSM total field inversion approaches. In addition, the algorith

​Modern Deep Learning Techniques Applied to Natural Language Processing by Authors

🔗 Modern Deep Learning Techniques Applied to Natural Language Processing by Authors

🎥 Intel: Leading the deep learning accelerator solutions
👁 1 раз ⏳ 1926 сек.
In this talk, we will discuss how Intel is leading the industry in developing AI accelerators and how new AI hardware like our Intel Nervana Neural Network Processors, paired with open-source software, will allow developers to harness the power of AI for any application.

Subscribe for more videos like this: http://hpe.to/6007Beguh

Visit our website: https://www.hpe.com

​Open-sourcing PyRobot to accelerate AI robotics research

🔗 Open-sourcing PyRobot to accelerate AI robotics research
Facebook AI is open-sourcing PyRobot, a lightweight, high-level interface that lets AI researchers get up and running with robotics experiments in just hours, with no specialized robotics expertise.

​How to Implement GAN Hacks to Train Stable Generative Adversarial Networks
https://machinelearningmastery.com/how-to-code-generative-adversarial-network-hacks/

🔗 How to Implement GAN Hacks to Train Stable Generative Adversarial Networks
Generative Adversarial Networks, or GANs, are challenging to train. This is because the architecture involves both a generator and a discriminator model that compete in a zero-sum game. It means that improvements to one model come at the cost of a degrading of performance in the other model. The result is a very unstable training …

​CNN Application-Detecting Car Exterior Damage(full implementable code)

🔗 CNN Application-Detecting Car Exterior Damage(full implementable code)
Recent advances in deep learning and computation infrastructure(cloud,GPUs etc.) have made computer vision applications leap forward: from…

🎥 Deep Q Learning w/ DQN - Reinforcement Learning p.5
👁 1 раз ⏳ 1940 сек.
Hello and welcome to the first video about Deep Q-Learning and Deep Q Networks, or DQNs. Deep Q Networks are the deep learning/neural network versi...

Здравствуйте, мне очень срочно нужен начинающий дизайнер. Разработка 3 баннера для фото в профиль магазина. Сразу говорю у меня нет бюджета.... понимаю,что многие меня пошлют,но всё ж надеюсь на Человека,который сможет помочь

​CNN Heat Maps: Saliency/Backpropagation

🔗 CNN Heat Maps: Saliency/Backpropagation
In this post I will describe the CNN visualization technique commonly referred to as “saliency mapping” or sometimes as “backpropagation”…

​How to Increase the Impact of Your Machine Learning Model

🔗 How to Increase the Impact of Your Machine Learning Model
Hint: Balance your efforts.

​The Uncanny Valley of Developing Technical Products

🔗 The Uncanny Valley of Developing Technical Products
In B2B software, one of the biggest selling objections is the “Build vs. Buy” argument. The objection is most prevalent at places with…

​Top 10 Great Sites with Free Data Sets

🔗 Top 10 Great Sites with Free Data Sets
Places to find free, interesting datasets and leverage insights from.

​Beating State of the Art by Tuning Baselines

🔗 Beating State of the Art by Tuning Baselines
Evaluating ML models by comparing to a baseline is popular, but how do we know if that baseline is as good as it could be?

​Submodular Batch Selection for Training Deep Neural Networks
https://arxiv.org/abs/1906.08771

🔗 Submodular Batch Selection for Training Deep Neural Networks
Mini-batch gradient descent based methods are the de facto algorithms for training neural network architectures today. We introduce a mini-batch selection strategy based on submodular function maximization. Our novel submodular formulation captures the informativeness of each sample and diversity of the whole subset. We design an efficient, greedy algorithm which can give high-quality solutions to this NP-hard combinatorial optimization problem. Our extensive experiments on standard datasets show that the deep models trained using the proposed batch selection strategy provide better generalization than Stochastic Gradient Descent as well as a popular baseline sampling strategy across different learning rates, batch sizes, and distance metrics.

🎥 Stanford CS230: Deep Learning | Lecture 5 - AI + Healthcare
👁 1 раз ⏳ 5178 сек.
Stanford CS230: Deep Learning | Lecture 5 - AI in Healthcare.

Andrew Ng, Adjunct Professor & Kian Katanforoosh, Lecturer - Stanford University htt...

​Создаем прототип для Sentiment Analysis с помощью Python и TextBlob

Что важно для команды разработчиков, которая только начинает строить систему, базирующуюся на машинном обучении? Архитектура, компоненты, возможности тестирования с помощью интеграционных и юнит тестов, сделать прототип и получить первые результаты. И далее к оценке трудоемкости, планированию разработки и реализации.

В этой статье речь пойдет как раз о прототипе. Который был создан через некоторое время после разговора с Product Manager: а почему бы нам не «пощупать» Machine Learning? В частности, NLP и Sentiment Analysis?
https://habr.com/ru/post/457168/

🔗 Создаем прототип для Sentiment Analysis с помощью Python и TextBlob
Что важно для команды разработчиков, которая только начинает строить систему, базирующуюся на машинном обучении? Архитектура, компоненты, возможности тестирован...

​Освобождаем руки нескольким аналитикам: API Livy для автоматизации типовых банковских задач
Наш телеграм канал - tglink.me/ai_machinelearning_big_data

Не секрет, что для оценки платежеспособности клиентов банки используют данные из различных источников (кредитное бюро, мобильные операторы и т.д.). Количество внешних партнёров может достигать нескольких десятков, а аналитиков в нашей команде наберётся лишь несколько человек. Возникает задача оптимизации работы небольшой команды и передачи рутинных задач вычислительным системам.

Как данные попадают в банк, и как команда аналитиков следит за этим процессом, разберём в данной статье.
https://habr.com/ru/company/homecredit/blog/457096/

🔗 Освобождаем руки нескольким аналитикам: API Livy для автоматизации типовых банковских задач
Привет, Хабр! Не секрет, что для оценки платежеспособности клиентов банки используют данные из различных источников (кредитное бюро, мобильные операторы и т.д.)...