Posts Tagged deep learning
Posted by Darshana V. Nadkarni, Ph.D. in Big Data -Cloud -IoT-Software -Mobile -Entrepreneurship, Biotech - Medical Device - Life Science - Healthcare on March 23, 2020
The looming threat of Covid-19 and the grim reality of the toll that novel coronavirus takes on humankind, makes it imperative that we find a way to prevent the spread faster and with less cost. Currently nine out of ten drug therapies fail mostly between phase 1 trials and regulatory approval. The estimated cost of developing a new treatment is around staggering US $2.6 billion. While this article won’t address it, I want to mention that many bio/pharma companies are working with repurposed drugs to find a cure and at least 69 drugs have been identified as treatment possibilities. Drugs also have side effects and need to be tested for safety. Let us focus for a while on search for the vaccine.
Many companies including Moderna, CureVac and BioNTech are working on vaccines. With the help of Machine Learning, Artificial Intelligence and other technologies the hunt for new pharmaceuticals and appropriate chemicals is expected to be quicker, cheaper and more effective. Novel coronavirus presents the most unprecedented challenge to date because of the speed with which it spreads.
Who better than a supercomputer made by a company that has been on the cutting edge of innovation for over a century, to take on this speed challenge? IBM scientists instructed the world’s fastest computer to tackle this challenge at its Oak Ridge National Laboratory in Tennessee. Summit can run 200 quadrillion computations per second. The scientists ran thousands of simulations to analyse which drug compounds could stop the coronavirus from infecting the host cells.
Summit, an IBM supercomputer equipped with the “brain of AI” identified 77 compounds (from over 8000 compounds) that could be efficacious in preventing coronavirus from spreading in the host. This is promising news in humanity’s quest for an effective vaccine against the virus. These findings are published in the journal ChemRxiv and give us hope although road is still long to get there.
With increasing computer processing power and advanced algorithms, AI has been employed to analyze large data sets with greater efficiency and will likely lead to many exciting innovations. While AI and ML show promise to change every industry sector for the better, artificial intelligence, machine learning and deep learning have become the most widely discussed topics in the healthcare sector and the excitement keeps growing.
I will be looking forward to hearing about new innovations AI/ML at #TiEcon2020. Stay tuned for new dates for the conference at http://www.tiecon.org and on this blog.
Posted by Darshana V. Nadkarni, Ph.D. in Big Data -Cloud -IoT-Software -Mobile -Entrepreneurship on April 13, 2018
Cloud revolution that began over a decade ago, has transformed business models and working practices around cybersecurity. And yet continuous innovation is imperative when it comes to cybersecurity. Cybercrime damages are expected to rise to $6 trillion annually, by 2021. Hence cybersecurity and cloudsecurity represent a rapidly growing market today with the potential for revenues in excess of $10 billion by 2022.
Increasing cybersecurity threats have encouraged organizations to leverage technology advances, and fight back with a variety of strategies and tactics. For instance, leveraging big data, enables a security solution provider to see bigger patterns and connect the dots to understand the threats. Similarly, deep learning, an advanced form of artificial intelligence, has made a huge impact on cyber security. When a machine learns what a malicious code looks like, it can identify unknown codes as benign or malicious with extremely high accuracy, and in real time. Other technology solutions that have been implemented include, running applications in containers (e.g. Docker), instead of virtual machines, advances in virtual customer premise equipment (vCPE), and more flexible, open and cloud-based WAN technologies rather than proprietary or specialized WAN technologies.
These and other strategies are aimed at identifying management systems, people and devices accessing the networks, verification of authorization levels, advanced threat analytics to flag behavioral changes, and virtualized security to follow and protect the data whereever it goes. Largest entrepreneurship conference, TiEInflect 2018, to take place in Santa Clara Convention Center, in May, will again feature focused Cybersecurity track with exciting keynotes and speakers from leading cloud and cyber security companies. Register for the conference at www.tieinflect.org .