The Who’s Who of Deep Learning – Part 1

Big Data analytics and deep learning can help organizations flourish, Be it in earning profits, improving operational systems or addressing the needs of captive markets. Deep learning gives big data platforms a human like cerebrum than can asses patterns and generate its own solutions. So, how did this technology come to be?

Several years back, tech geniuses already recognized the impact of big data on businesses and, industries, and how it can make information analyses simpler and faster to execute. Through heavy research, testing, and timely collaborations, they helped turn bulks of random data into ‘insider’ information on prospective customers, the quantified metrics of employee performance, and a lot more. Let’s take a closer look at these personalities and their contributions to deep learning.

Geoffrey Hinton: The Man Who Tried to Mimic the Human Brain

A typical college kid often thought  about cars, fashion trends, and partying. For a young Geoffrey Hinton,it was different. He was fascinated about  the human brain and its complexities. A friend introduced the idea of the brain acting like a hologram. Memories are stored through an expansive web of neurons, rather than being kept in one location, similar to the way a series of light rays form a holographic image.

This fascination would later extend to Hinton’s career. After taking up experimental psychology and earning a doctorate degree in artificial intelligence, he eventually founded the Neural Computation and Adaptive Perception (NCAP) program. Through this program,  it gave him a chance to team up with esteemed researchers and scientists like Yann Lecun and Yoshua Bengio.

Hinton’s idea of mimicking organic intelligence with the use of data and computers was clearly ahead of its time. In the 1980s, the technology at hand was incapable of bringing Hinton’s concepts to life, thus was largely ignored by the AI community. Undeterred, Hinton and his NCAP team went on to develop the backpropagation algorithm for training neural nets.

Decades elapsed and eventually the technology caught up with the Hinton’s requisites. In the mid-2000s, larger datasets were incorporated and more effective algorithms were developed, eventually winning them AI competitions all across the globe. It didn’t take long before tech giant Google took notice of Hinton’s work and hired him to use his expertise  in neural networks. This  paved the way for deep learning adoption across industries.

As it turns out, Hinton was right all along, and now he is leading the charge in continually innovating data analytics. Not bad for something that started out as a juvenile fascination.

Andrew Ng: Empowering People with Education and “Deep Learning”

Coursera is currently the largest source free online education covering a multitude of courses. Remarkably, its founding came at the heels of a Machine Learning class, which introduced and dispensed concepts on big data, analytics, and deep learning. Coursera co-founder, Andrew Ng, happened to be the instructor, and deep learning was his key to success.

Outside of teaching, Ng worked with Google and spearheaded a deep learning project called Google Brain, where he worked with tech visionary Geoffrey Hinton. This led to the development of algorithms for massive-scale deep learning, which increased  the scope and functionality of big data analytics. Ng and his crew, in testing the algorithms, came up with “Google cat” – it gained fame for detecting untagged cat videos on YouTube.

Apart from Google and Coursera, Ng works as the chief scientist of Baidu, a large search engine based in China. Baidu followed Google’s lead and also applied deep learning techniques to its operations. Ng’s efforts are continually paving the way for new data solutions. Now, this driven data expert is working on innovating voice recognition, computer vision, autonomous driving powered by deep learning, and he has no plans of slowing down.

Yann LeCun: The Tagging Savant

Ever wondered how Facebook manages to recognize a person’s facial features? Upon uploading pictures, Facebook – through deep learning – is able to tag users to the right faces. This is the handiwork of respected data researcherYann LeCun, whose body of work contained research and development projects with large multinational companies.

As a young professional in the 1980s, LeCun worked as an associate researcher for George Hinton’s group, NCAP. There, he helped developed early forms of deep learning algorithms which attracted the attention of a number of successful tech conglomerates.

In 1988, LeCun joined AT&T where he applied his knowledge on deep learning and developed an image recognition feature called Convolutional Neural Networks. Additionally, he used machine learning to create a system that recognizes the validity of bank checks, which gained widespread use throughout the US. LeCun developed it further and married it with an application that recognized a person’s handwriting. This system took into account the common strokes used and the boldness of the impressions.

Because tech geniuses aren’t satisfied with their laurels, LeCun helped improve the image processing system of DjVu, a file format for compressing scanned documents which he co-founded. He also co-developed a programming language called Lush.

With a number of achievements tucked under his belt,  it’s no wonder that the world’s biggest social networking site, Facebook, took interest in LeCun.. He was named Facebook’s Director of AI Research in December 2013, where he developed the image tagging feature that is still being applied and enhanced to this date. For his pioneering work and his overall effort as a scientist for deep learning, Lecun received the IEEE Neural Network Pioneer Award.

Yann LeCun has not only  helped tag our Facebook photos, but he made bank checks more secure and furthered the science of  image processing. Now we’re tagging him as one of the most admirable figures in the field of machine learning, data analytics and deep learning.

Through the contributions of George Hinton, Andrew Ng and Yann LeCun, we have come to utilize and appreciate the immeasurable conveniences given  to us by modern technology. They have revolutionized the way people do business and how an organization properly serves its targeted demographics. Remarkably, the use of applied data analytics is relatively young; there is still a lot of  room for new solutions to be developed. Hinton, Ng, LeCun might already be working on them, for all we know.

Related Posts
The Who’s Who of Deep Learning – Part Two