How to build an entire ML pipeline, including data transformation and model training, without code.
Category: robotics/AI – Page 2,137
Tesla is adding remarkable detail in its Full Self-Driving Visualizations
Most of the pieces of Elon Musk’s Master Plan, Part Deux are already in place. Tesla’s mass-market cars, the Model 3 and Model Y, have already been released. The Solar Roof is finally seeing a ramp. And the release of a feature-complete version of the company’s Full Self-Driving suite seems to be drawing closer.
For Tesla’s Full Self-Driving suite to be feature-complete, the electric car maker would need to master inner-city driving. FSD already works for highway driving with Navigate on Autopilot with automatic lane changes. But when it comes to inner-city streets, Full Self-Driving still has some ways to go. Fortunately, if Tesla’s v10.2 2020.12.5 release is any indication, it appears that more and more aspects of city driving are becoming recognized by the company’s neural networks.
Automatic diagnosis of the 12-lead ECG using a deep neural network
The role of automatic electrocardiogram (ECG) analysis in clinical practice is limited by the accuracy of existing models. Deep Neural Networks (DNNs) are models composed of stacked transformations that learn tasks by examples. This technology has recently achieved striking success in a variety of task and there are great expectations on how it might improve clinical practice. Here we present a DNN model trained in a dataset with more than 2 million labeled exams analyzed by the Telehealth Network of Minas Gerais and collected under the scope of the CODE (Clinical Outcomes in Digital Electrocardiology) study. The DNN outperform cardiology resident medical doctors in recognizing 6 types of abnormalities in 12-lead ECG recordings, with F1 scores above 80% and specificity over 99%. These results indicate ECG analysis based on DNNs, previously studied in a single-lead setup, generalizes well to 12-lead exams, taking the technology closer to the standard clinical practice.
The Ancient History of Artificial Intelligence
The Ancient History of AI:
This video was made possible by Surfshark. Sign up with this link and enter the promo code ‘Futurology’ to get 83% off and an extra month free of your VPN plan! https://surfshark.deals/Futurology
Artificial intelligence has been a topic of growing prominence in the media and mainstream culture since 2015, as well as in the investment world, with start-ups that even mention the word in their business model, gaining massive amounts of funding.
While to many, the hype around AI may appear sudden, the concepts of modern artificial intelligence have been around for over a century and extending further, the concept of artificial intelligence and artificial beings have been in the minds of humans for thousands of years.
To better understand and appreciate this technology and those who brought it to us as well as to gain insight into where it will take us: sit back, relax and join me in an exploration on the history of artificial intelligence.
Israeli researchers at Hebrew U develop faster, cheaper COVID-19 test
Researchers at the Hebrew University of Jerusalem announced on Sunday that they have developed a new method of testing for COVID-19 which is not only 4–10 times faster than the tests most commonly used today, but also significantly cheaper, while supplying the same level of accuracy. Moreover, most of the materials required to perform the new test are already available in Israel, easing significantly both the country’s dire shortage of testing materials and its heavy economic dependence on foreign commercial markets. The method was developed in the labs of Prof. Nir Friedman of the Institute of Life Sciences and the School of Engineering and Computer Sciences and Dr. Naomi Haviv of Hebrew University’s Neuroscience Research Center, and is based on materials which are not affected by global shortages and can be recycled for repeated used on future tests. The method commonly used today for COVID-19 testing involves extracting RNA molecules from a patient’s sample to determine if the molecules produced have viral RNA within them, which confirms the presence of the coronavirus. The new test developed by the researchers performs the same action, but is made from more commonly attainable materials, that produce results at a much higher speed. Dr. Naomi Haviv said that “We have an efficient RNA extraction method, 4–10 times faster than the current method. It is based on magnetic beads and can be performed both robotically and manually.”
Other than the magnetic beads, all of the other materials needed to perform the tests are available for purchase in Israel. The beads themselves are recyclable and can be reused to perform future tests. “The robotic method has already undergone a series of tests at Hadassah Hospital, using hundreds of samples from patients — and is now becoming operational.”
The new test would supply results 4–10 times faster than the current method, and is made almost entirely from materials which are easily available for purchase within the country.
Programming a Quantum Computer
It is a common misconception that quantum computers are not yet ready for applications and the technology still has many years before becoming useful. In this article we will take a look at some of the basic principles of programming a quantum computer and address this misconception. We will look at free, open-source software such as QISKit from IBM, as well as the Quantum Machine Learning software PennyLane. We will also explain how you can run your programs on actual quantum computers in the cloud at IBM. In a follow-up article we will talk about some applications in machine learning that are ready for use currently to anyone with a bit of curiosity.
Using artificial intelligence to search for new exotic particles
Nowadays, artificial neural networks have an impact on many areas of our day-to-day lives. They are used for a wide variety of complex tasks, such as driving cars, performing speech recognition (for example, Siri, Cortana, Alexa), suggesting shopping items and trends, or improving visual effects in movies (e.g., animated characters such as Thanos from the movie Infinity War by Marvel).
Traditionally, algorithms are handcrafted to solve complex tasks. This requires experts to spend a significant amount of time to identify the optimal strategies for various situations. Artificial neural networks — inspired by interconnected neurons in the brain — can automatically learn from data a close-to-optimal solution for the given objective. Often, the automated learning or “training” required to obtain these solutions is “supervised” through the use of supplementary information provided by an expert. Other approaches are “unsupervised” and can identify patterns in the data. The mathematical theory behind artificial neural networks has evolved over several decades, yet only recently have we developed our understanding of how to train them efficiently. The required calculations are very similar to those performed by standard video graphics cards (that contain a graphics processing unit or GPU) when rendering three-dimensional scenes in video games.
SpaceX developed new machines to speed up the construction of Starship prototypes
Featured Image Source: SpaceX
SpaceX is in the process of developing its next-generation spacecraft. It will be a two-stage launch vehicle, consisting of Starship (a spaceship), and a rocket booster called Super Heavy. Starship will be capable of conducting voyages to the moon and Mars. Super Heavy will only be needed to take Starship out of Earth’s atmosphere. It will be capable of returning from space in order to be reused. SpaceX has developed some of the world’s most technologically advanced rockets. The Falcon 9 rocket’s first-stage booster is capable of being launched, and return to land vertically on autonomous drone ships at sea in order to be reused again. No one in the rocket industry has achieved that level of reusability. Making Starship reusable is key towards decreasing the cost of spaceflight to make life multi-planetary.
Starship’s development is partially funded by Yusaku Maezawa, a fashion entrepreneur who dreams of going to the moon. Maezawa booked Starship’s first crewed flight, it will be a circumlunar voyage scheduled for the year 2023. SpaceX has an ambitious deadline to meet. The rocket company is manufacturing a production line of Starships at its South Texas assembly facility located in Boca Chica Beach, Brownsville, TX. Elon Musk, founder and chief engineer at SpaceX, says a high iteration rate is needed to make rapid progress in the development of Starship. To rapidly innovate, many vehicles must be manufactured and tested rapidly. Therefore, the company aims to ramp up its production to build one Starship per week. “Production is at least 1,000 percent harder than making one of something. At least 1,000 percent harder,” Musk said. This year, the company manufactured three stainless steel prototypes of the craft, two imploded during pressurization tests.