Lifeboat Foundation

Big-data analysis consists of searching for buried patterns that have some kind of predictive power. But choosing which “features” of the data to analyze usually requires some human intuition. In a database containing, say, the beginning and end dates of various sales promotions and weekly profits, the crucial data may not be the dates themselves but the spans between them, or not the total profits but the averages across those spans.

MIT researchers aim to take the human element out of big-data analysis, with a new system that not only searches for patterns but designs the feature set, too. To test the first prototype of their system, they enrolled it in three data science competitions, in which it competed against human teams to find predictive patterns in unfamiliar data sets. Of the 906 teams participating in the three competitions, the researchers’ “Data Science Machine” finished ahead of 615.

In two of the three competitions, the predictions made by the Data Science Machine were 94 percent and 96 percent as accurate as the winning submissions. In the third, the figure was a more modest 87 percent. But where the teams of humans typically labored over their prediction algorithms for months, the Data Science Machine took somewhere between two and 12 hours to produce each of its entries.

Read more

Madrid, Spain (Scicasts) — A novel way to count white blood cells without a blood test, simply by applying a small device on the fingertip, is being developed by a team of young bioengineers.

The technology, that combines an optical sensor with algorithms, has already three prototypes on the go and is specially designed to be used on chemotherapy patients, who could know their immune system levels in real time. It could also serve to detect serious infections.

A group of young bioengineers from various countries, including Spaniard Carlos Castro, is developing a portable device capable of counting white blood cells in real time, without requiring a blood test. The system includes an innovative optics sensor through the skin that can observe white cells as they flow past a miniature lens. This new device could be applied to improve the treatment of patients who are left immunosuppressed after chemotherapy treatments and to prevent sepsis.

Read more

The Blue Brain Project is a vast effort by 82 scientists worldwide to digitally recreate the human brain. While still far from that goal, the team revealed a breakthrough that has already provided insight into sleep, memory and neurological disorders. They created a simulation of a third of a cubic millimeter of a rat’s brain. While that might not sound like much, it involves 30,000 neurons and 37 million synapses. In addition, the simulated level of biological accuracy is far beyond anything so far. It allowed them to reproduce known brain activities — such as how neurons respond to touch — and has already yielded discoveries about the brain that were impossible to get biologically.

To create the simulation, researchers did thousands of experiments on rat brains over a 20 year period, logging each type of synapse and neuron discovered. That led to a set of fundamental rules describing how neurons connect to synapses and form microcircuits. Using the data, they developed an algorithm to pinpoint the synapse locations, simulating the circuitry of a rat’s brain. All of that data was then run through a supercomputer: “It was only with this kind of infrastructure that we could solve the billions of equations needed,” said software lead Felix Schurmann.

Read more

Adobe engineer Ashutosh Jagdish Sharma demonstrated the technology on stage, enlisting the help of host Kim Chambers and Parks and Rec star Nick Offerman to act as stand-in tourists who were getting in the way of the desired shot. When the smartphone was held in place, Monument Mode was able to slowly erase the “tourists” from the image, building up a clear version of the photo slowly as human obstructions moved around. Even though Chambers and Offerman remained inside the frame, the final result showed the background only, the feature able to create a clear image from multiple shots.

Traditionally photographers have been able to remove tourists and other obstructions after their photos are taken with clever Photoshop work, by taking multiple shots, or by taking them from various angles. But Monument Mode works in real-time, cutting down on legwork, and requiring fewer photo-editing skills. The company says it the feature ”uses a new algorithm to distinguish moving objects from fixed ones,” but notes that it’s still only a tech preview, and that it may not come to fruition. That said, the company has a history of swiftly incorporating technology shown off at its MAX conferences. Adobe first detailed its “dehaze” feature during the same segment at last year’s show — it now comes as standard in Lightroom.

Read more

The future of Eco conservation?

Deforestation downs 10 billion trees around the globe annually. Replanting trees by hand is slow, expensive, and barely puts a dent in reversing the damage. But one startup wants to use drones that can reforest our increasingly tree-strapped Earth, on a big enough scale to replace slow and expensive hired humans.

The small company, called BioCarbon Engineering, says unmanned aerial vehicles are a great way of covering ravaged woodlands with seedlings that can repopulate the area’s tree population. Around the world, forests and jungles are still being leveled due to lumber overproduction, strip surface mining, urban expansion, and land use for agriculture.

Continue reading “This Startup Wants To Plant One Billion Trees a Year Using Drones” »

Memory loss is a truly devastating part of dementia, but this invention aims to fix that by bypassing the damage, and repairing long term memory.

Alzheimer’s and dementia are complex diseases, and there’s currently no effective treatment. Given the unpleasant nature of the disease, there’s an urgent need for results. Instead of taking the usual biological route, one team has constructed a prosthetic made up of a small electrode array — which can help re-encode short term memory into long term.

Built using decades of research, the device operates using a new algorithm based on accumulated neural data. New sensory information is normally translated into a quick memory and transported as an electrical signal through the hippocampus, potentially for long term storage. If this region is damaged then the process is disturbed, and new experiences fail to be encoded. Alzheimer’s patients can often remember childhood events, but struggle with recent experiences; specifically because of this hippocampal damage.

Read more

Researchers have developed a computer algorithm that mimics the brain’s electrical signalling and helps memory. The FT reports.

Read more

The word on every tech executive’s mouth today is data. Curse or blessing, there’s so much data lying around – with about 2.5 quintillion bytes of data added each day – that it’s become increasingly difficult to make sense of it in a meaningful way. There’s a solution to the big data problem, though: machine learning algorithms that get fed countless variables and spot patterns otherwise oblivious to humans. Researchers have already made use of machine learning to solve challenges in medicine, cosmology and, most recently, crime. Tech giant Hitachi, for instance, developed a machine learning interface reminiscent of Philip K. Dick’s Minority Report that can predict when, where and possibly who might commit a crime before it happens.

Read more

Turing Test-style competitions in writing stories, music & poetry.

HT: @Grady_Booch

DigiLit is a competition that encourages the creation of algorithms able to produce a “human-level” short story of the kind that might be intended for a short story collection produced in a well-regarded MfA program or a piece for The New Yorker. The prize seeks to reward algorithms that could, for example, write stories for a creative writing class in which students are asked to submit a new short story each day. (Artwork by Annelise Capo http://www.annelisecapossela.com)

Read more