Add idea


«    September 2019    »










DateDate: 14-09-2019, 06:16

Scientists from the United States have developed an algorithm based on artificial intelligence, which calculates when an employee is about to quit. The program will help company executives retain an old employee or start looking for a new employee
Professors Brooks Holt and David Allen used Big Data and machine learning algorithms to develop a technology that defines indicators that talk about employee plans to quit.
The first indicator is a shocking frame rate. Usually this is a change of management or a major acquisition that affected the value of the shares, received media coverage or attracted the attention of lawyers.
The second indicator is the employee’s relationship with the organization in which he works. Researchers relied on open data, such as the number of previous jobs, tenure at the last company, skills, education, gender, and geography. According to a three-month study, it turned out that about 64% of workers want to change jobs, CNBC reports.
Replacing a qualified employee can cost a company significant financial losses, given the time spent, HR work, and newcomer training, Holt said. Some companies that cannot find a candidate for an open position lose about $ 1 million annually.
Scientists are confident that the technology will help experienced managers and employers learn more about what their employees really value in their work, as well as keep them and create truly comfortable conditions, and not just give empty promises.

DateDate: 13-09-2019, 05:49

It's no secret that despite its pretty good performance, absolutely all lithium-ion batteries have a certain degree of danger - however, given their performance, many people simply prefer not to pay attention to it. However, a talented team of chemical engineering specialists from the University of Houston in the USA decided to do their own thing and present something safer and no less productive in their alternative - because the scientists created a new type of organic cathode with a high degree of flexibility and adaptability to the interface with solid electrolyte - which generally increases the energy efficiency of such a battery.
Thus, scientists, in fact, managed to invent a sodium-ion battery, which has not only a higher degree of conductivity and charge capacity, but which also boasts a significantly higher degree of safety even with prolonged and intensive use. This is because a new organic cathode called PTO - pyrine-4,5,9,10-tetraone - offers truly unique opportunities for increasing energy efficiency, mainly due to a thinner layer of pressure against the resistive interface between the cathode and electrode.
Of course, this development, despite its high level of performance and safety, still can not fully compete with traditional lithium-ion batteries, because at first it must be compared with them in long-term power consumption. In addition, the development is distinguished by a rather controversial life cycle of the organic cathode itself.
In everything else, it becomes clear that such a sodium-ion battery can become the future of absolutely all types of electronics - moreover, subsequent modifications will probably be somehow connected with the need to increase the interface gasket between the cathode and electrode. Scientists are in no hurry to cover all aspects and details regarding the new battery, but now we can assume that they will soon please us again!

DateDate: 11-09-2019, 05:48

The neural network was taught to transform the image into music.
Researchers from the University of Amsterdam presented a neural network that can relate visual sensations to sounds - and turn pictures into music. The work was published in an article on, and will be presented at the ICCVW 2019 conference.
When teaching the algorithm, the researchers did not show him how to correlate the image with the music - the neural network received these skills on its own using the teaching method without a teacher.
The algorithm is built on the principle of an auto-encoder and consists of an encoder and a decoder that work with different data - in this case, with images and sounds.
The encoder studied on the MNIST dataset, which contains 60 thousand handwritten characters, and on the Behance Artistic Media dataset, which includes about 180 thousand oil and watercolor paintings.
The neural network first performs the conversion from the image to music, it performs the reverse conversion from the received music to a new image, after which it is compared with the original one.

DateDate: 10-09-2019, 05:53

Fairphone 3 has a less harmful environmental impact
The Dutch startup Fairphone presented a new “ethical” smartphone, the components of which can be improved by yourself, TechCrunch reports.
Fairphone 3 production has a less harmful effect on the environment and better working conditions, says the manufacturer.
The smartphone runs on Android. The model has a 5.7-inch display, Qualcomm Snapdragon 632 processor, 4GB of RAM, 64 GB of memory, two cameras - 12 megapixels and 8 megapixels, a fingerprint scanner, two slots for nano-SIM, a 3000 mAh battery and an NFC system. Its price is € 450.
The main difference between Fairphone 3 is the environmental friendliness of production, which has less negative impact on the environment. Another of its features is the ability to replace obsolete or damaged modules, instead of throwing away a smartphone and buying a new one. This will reduce both personal expenses and the proportion of toxic waste.
Included with the phone is a screwdriver for the manual replacement of smartphone modules, such as a display, camera and the like.
As TechCrunch notes, despite all its innovation, the Fairphone 3 is "nauseous plain." And it is this particularity of the gadget that can become a challenge for electronics, which is not distinguished by ethics and concern for the environment.

DateDate: 9-09-2019, 05:56

Artificial intelligence was used to analyze digital X-ray images of the Ghent Altar, which is located in the Catholic Cathedral of St. Bavon in Belgium. Researchers note that AI findings will help improve understanding of the details of art masterpieces and will open up new possibilities for the study and preservation of artwork.
By analyzing complex x-ray images, the new algorithm allows art critics, museum keepers and scientists to better understand the paintings, and the information received can help specialists protect and restore fragile objects.
X-ray images are a valuable tool for studying and restoring paintings, as they can help establish the state of the work and give an idea of the artist’s technique, the researchers noted.
“We would like to see how the development of similar approaches focused on artificial intelligence will affect our ability to detect hidden features in the picture that people have not noticed,” they added.
Helen Dubois, project manager for the preservation of the Ghent Altar, noted that “using artificial intelligence to process X-ray images will provide very useful tools for decoding complex technical images. The structural defects of wooden supports, as well as layers of soil and paint can now be diagnosed with greater accuracy.