RSS Feed
April 18th, 2016 
Awake Goy 
Susanne Posel (OC) : The Massachusetts Institute of Technology (MIT) Computer Science and Artificial Intelligence Lab (CSAIL) have developed a way to predict hacker attacks with an 86% accuracy using artificial intelligence (AI).
Kalyan Veeramachaneni, lead researcher for the project said: “You have to bring some contextual information to it. On day one, when we deploy the system, it’s [only] as good as anyone else.”
Affectionately called AI2, this robotic intelligence combines the awareness of anomalies in internet traffic with the eloquence of human input to decipher whether a cyberattack is eminent.
AI2 is described as a “secret weapon” that “fuses together three different unsupervised-learning methods, and then shows the top events to analysts for them to label.”
Journalist Jamie Condliffe explained : “With that feedback, it takes on board whether or not it should be classifying the events as attacks or not, then refines its internal models. Over time, it becomes better able to differentiate signal from noise, showing fewer incorrect results and in turn saving the expert’s time.”
By using predictive models, AI2 is able to “prevent attacks [that] might happen the following day” which would allow for corporations to “bolster security by… requiring additional confirmation from customers.”
Veeramachaneni added: “You can think about the system as a virtual analyst. It continuously generates new models that it can refine in as little as a few hours, meaning it can improve its detection rates significantly and rapidly.”
MIT is no stranger to the advancement of AI. Just last year they developed an algorithm that is expected to enhance domestic robots with more precise abilities in identifying “objects in cluttered environments” such as the interior of a home.
The CSAIL team created “a system using an off-the-shelf algorithm to aggregate different perspectives can recognize four times as many objects as one that uses a single perspective, while reducing the number of misidentifications.”
According to the report, the algorithm used by the team aggregated “different perspectives” and was able to improve the robot’s ability to recognize “4 times as many objects as one deploys [in] a single perspective.”
Other experiments into robotics by MIT includes:
• Flat-pack robots that can fold up, unfold and walk away all by itself
• Jibo a robot built to work alongside humans, affectionately designed after Wall-E’s “EVE”
• An Iron Man suit
• Small robot cubes that have no external moving parts called M-Blocks robots (MBR)
• Technology the to use for Wi-Fi that emits a “signal to track moving humans – even behind walls” called Wi-Vi
MIT’s discovery would help improve domestic robots such as Budgee who is tasked with assisting human in need and the elderly by becoming “part of the family unit” and integrating into “our society and mindset”.
Science fiction is becoming a reality with the introduction of service robots that is expected to “influence human-robot interaction”.
These robots will be autonomous, mobile and perform functions humans may not want to do in the future.
Susanne Posel, Occupy Corporatism
Source Article from http://nsnbc.me/2016/04/18/mit-creates-breakthrough-ai-that-can-detect-85-of-cyberattacks/
Related posts:
Views: 0
Posted in 
Tags: 
