Machine-Learning

NVIDIA CLARA DISCOVERY Eco Nvidia’s entry into the medical field. From the cover we can see a ring structure: Natural Language -> Viral Gene Sequence -> Viral Protein Structure -> New Drug Docking with Viral Protein -> Molecular Docking Dynamics Simulation -> Clinical Trial of Drug Candidate Detecting Medical Images of Patient’s Lungs -> Natural Language Introduction At the Nvidia GTC 2020 Fall Conference, Nvidia announced the NVIDIA CLARA DISCOVERY drug development framework, which aims to accelerate the progress of global drug development. With the news of Nvidia’s acquisition of ARM before, I realized that Nvidia has the ambition to rule the edge computing field in the future. Not surprisingly, this GTC conference Nvidia updated many product lines, including large to data center computing module BlueField-2, small to edge field jetson Nano 2G, the most exciting for me is still ** CLARA DISCOVERY**, because my own short-term goal is how to use edge computing in the medical field, such as individuals can calculate the current status of their own diseases through edge computing devices combined with their own diagnostic data obtained from hospital examinations, as well as judgmental detection of Covid-19 CT, ChestXRay. ...

Considering the existence of a large number of asymptomatic people in this epidemic, so this November holiday myself decided to DeepMind 2020 machine learning of the remaining six courses to finish, temporarily do not consider going on a trip. Course address: deep learning lecture series 2020 There is also a course that I have been following HarvardX’s Tiny Machine Learning (TinyML) program harvardx-tiny-machine-learning?utm_campaign=Professional%20Certificates%20%20&utm_medium=email&_hsmi=96130954&hsenc=p2ANqtz– 7LiwHUWYNB_CyU36B4PMjUk1cf7LS8UNQHZKmoUGBtZ0LctZw691q9Ixa5gc5vwEeOVBkGjCGF1vAOA8hr_Jb_akwxw&utm_content=95985108&utm_source=hs email), that is, the course is a little expensive $ 537, and then see for yourself whether there is a discount behind, because their future development direction tends to TinyML, so I will always pay attention to the latest developments in this field, if there are friends who want to buy this course together, you can contact me by email [email protected]. ...

Using graph neural networks in high-energy particle physics Reference image and article address: [The next big thing: the use of graph neural networks to discover particles](https://news.fnal.gov/2020/09/the-next-big-thing- the-use-of-graph-neural-networks-to-discover-particles/) Reflections The article describes the development of a set of graph neural network models by the Institute of High Energy Physics Fermilab in order to filter valuable images from the huge amount of particle collision images, which are deployed in the LHC lab to directly process the data generated during particle collisions. ...

Sidebar Behind why deep learning is so powerful, we can think of the human brain’s neurons, which individually can’t do anything, but a bunch of neurons together are immensely powerful. Watch a movie about how cells and proteins work in living things, and be amazed at how tiny microscopic worlds can work in such subtle ways. Inner Life Of A Cell - Full Version Sometimes when I think about the above questions, I get the urge to learn more about how the world works, whether human communication can be likened to molecular interactions, why a bunch of >neurons put together can create intelligent life forms, how the free consciousness of such life forms is formed, how the transfer of information between neurons is encoded, and what is happening at a deeper level. Whether the field of study can continue to exploit certain features of the brain > neurons to enhance the design of models. ...

A Review of Automated Diagnosis of COVID-19 Based on Scanning Images In the last week, I have been optimizing the Covid-Net I developed with my thesis and other research materials to diagnose the deep learning model of Covid19, and the accuracy I got from training on 20,000+ xRay datasets is less than 90%. When the model became more and more complex, the video memory was obviously not enough, so I sold the DJI Inspire 2 and used the money I earned to buy a TITAN RTX 24G, and then continued to optimize the model and increase training. of the batch_size. ...