Machine learning and AI neural networks - The EE

Machine learning and AI neural networks

Santhosh Srirambhatla of Blue Ridge

Artificial intelligence is already part of our everyday lives. It is also the future. The easiest way to think of their relationship, the artificial intelligence machine learning and deep learning, is to visualise them as concentric circles.

With AI, the idea that came first, the largest machine learning, which blossomed later. And finally, deep learning, which is driving today’s AI explosion, fitting inside both, says Santhosh Srirambhatla, chief technology officer of Blue Ridge.

AI has been part of our imaginations and simmering in research labs since a handful of computer scientists rallied around the term at the Dartmouth Conferences in 1956 and birthed the field of AI. In the decades since, AI has alternately been heralded as the key to our civilisations brightest future and tossed on technology’s trash heap as harebrained notion of over-reaching propeller heads.

Over the past few years, AI has exploded. Much of that has to do with the wide availability of GPUs that make parallel processing ever faster, cheaper, and more powerful. It also has to do with the simultaneous one-two punch of practically infinite storage and a flood of data of every stripe, images, text, transaction, and mapping data.

AI is quite broad in scope. And machine learning, which is a part of AI, is a small part of it. Right now, you can consider what we do in AI as a very narrow form of AI. Technologies that are able to perform specific tasks as well as, or better than humans, can be done using Narrow AI. They are very specific, which means there is a lot more to come in the field of AI.

AI, broadly speaking, is the science of making machine smart.

The systems implemented today are a form of Narrow AI, a system that can just do a few well-defined things such as recognising objects and gestures. It can do them as well as humans. Most AI systems involve some sort of machine learning or a deep learning technique, it also uses other technologies such as natural language processing, speech recognition, sound detection, and sentiment analysis.

Machine learning is a specific type of AI

Machine learning is really a specific methodology within the vast scope of AI. It is all about feeding computers data and teaching them to identify patterns. This is different from computer programming where specific business rules or logic are defined to perform a certain task.

Machine learning passes data and allows the computer to decide the patterns from the data. Once the model is trained, that model is used to make predictions for the data not seen previously.

Machine Learning users do not want to include too many features also because that results in over-fitting of the model and reduces the accuracy of the model for future predictions. Identifying and keeping the right features comes from experience; this is the purview of data scientists.

Machine learning capabilities transforms pricing optimisation

Machine Learning capabilities, such as those used by Blue Ridge in a price optimisation product, continually optimise across all products (not just top-tier items). Segment and perfect pricing for each customer gives distributors the upper-hand in supplier negotiations.

Comprehensive pricing strategies simulate different pricing scenarios and predict the impact of a price change before implementing it. Similarly, competitive positioning allows for an immediate response to price changes from competitors, as well as supplier rules, segmentation, positioning, and price change frequency.

Customer segmentation identifies customer performance, discounting, and pricing opportunities based on willingness-to-pay and past performance. These price optimisation data combined with suite of Supply Chain Planning solutions, help increase forecast accuracy, improve customer satisfaction, and assure product availability to customers without creating a costly inventory surplus.

Neural network forecasting

Neural Network forecasting is basically a deep learning technique. It uses demand history as input to determine the forecast. The simulations run will increase the forecast accuracy for fast moving items.

The author is Santhosh Srirambhatla, chief technology officer of Blue Ridge.

About the author

Santhosh Srirambhatla brings more than 20 years of experience in software development, consulting services, customer success, and information technology. He is responsible for all aspects of product development, including cloud operations.

Prior to joining Blue Ridge, Santhosh led successful implementation of supply chain solutions and transformations for enterprise organisations. His background also includes extensive S&OP, supply chain planning and logistics experience with management and consulting roles at Steelwedge, Honeywell and i2 Technologies. Srirambhatla holds a Master of Science in Industrial Engineering from Arizona State University and a Bachelor of Engineering in Mechanical Engineering from Osmania University.

Follow us and Comment on Twitter @TheEE_io

By continuing to use the site, you agree to the use of cookies. more information

The cookie settings on this website are set to "allow cookies" to give you the best browsing experience possible. If you continue to use this website without changing your cookie settings or you click "Accept" below then you are consenting to this.