Security solutions that make use of artificial intelligence, neural networks and deep learning are on the rise. Solution providers are increasingly trying to take advantage of these novel technological developments to offer more advanced, secure and safer products.
Security solutions that make use of artificial intelligence, neural networks and deep learning are on the rise. Solution providers are increasingly trying to take advantage of these novel technological developments to offer more advanced, secure and safer products.
However, despite this, not many security professionals understand the different concepts that machine learning and deep learning represent. It is in this context that a recent webinar hosted by Intel on Artificial Intelligence becomes relevant.
“We define AI as a technology that uses deep natural language processing and ability to understand and address questions and provide recommendations and directions,” said Chwee Chua, Big Data and Analytics and Cognitive Computing, IDC APAC at the webinar.
“Such system hypothesizes and formulates possible answers based on available evidence and can be trained to manage vast amounts of content. And it automatically adapts and learns from its mistakes and failures.”
“One of the questions that we often get is ‘what is the difference between machine learning and deep learning?’” Chua continued, adding that in general, machine learning is a subset of AI encompassing a range of algorithms to enable a trend or pattern recognition over time. It can be supervised, i.e. with expert training, or unsupervised, with no inputs from humans
A subset of machine learning is deep learning. Deep learning has many applications in today's world. From speech recognition to image recognition to even biomedical informatics. In general, one can think of it as a cascade, many layers of nonlinear processing unit for feature extraction and transformation. Each successive layer, uses the output from the previous layer as input.
Basically neural networks are loosely designed based on the biology of our brain. It’s simulating how we humans think. All this interconnects between the neurons, where the neurons can connect to any other neurons with certain physical distance. So artificial neural networks have several layers of connections and directions of data propagations.
Their applications include pattern analysis, classification etc. which are all based on the learning of multiple layers of features or representations of data. The higher level features are divided from the lower layer features to form a hierarchical representations.
"Think of how a baby learns,” Chua said. “How does a baby learn to recognize a cat? First, a baby identifies that it is an animal, and then it recognizes it has four legs, followed by it has whiskers, or color before finally learning to differentiate whether it is a cat or a dog. Similarly this is how deep learning kind of trains the system by going through several layers of analysis.”