Builders: (Bring a Ubuntu laptop with ROS installed if you can) 

• Tom Sommerville
• Edward Metzler
• Edward Epp

• Michael Brinkley: e1ectr0n (https://www.e1ectr0n.com)

Come to join us in our Bowling Night. You will have the opportunity to meet engineers from all fields. 

Talk up and throw some bowling balls! 

Highlights: During this presentation we will take a look "under the hood" of the mathematics which is driving all the excitement. Advanced math concepts are not required. A demonstration will be done with a Python coded example.

The real breakthrough for the modern Artificial Intelligence (A.I.) and Machine Learning (M.L.) technology explosions started back in 1943 when researchers McCulloch & Pitts came up with a mathematical model to represent that function of the biological neuron; nature’s gift that allows all life to operate and learn over time. Eventually this research would give birth to the A.I. revolution that we are experiencing today!

Before there can be any discussions of deep learning, there are some fundamental concepts and terms that need to be understood which begins with the introduction of the perceptron and its' structure. The perceptron is the basic building block of Artificial Neural Networks (ANN). While there are other approaches to A.I., which the other approaches will be briefly discussed, the ANN is the most fundamental approach and more closely relates to the solution to learning that nature has provided all of life; as we know it. There are also different classifications of ANNs, and they will be discussed as well, but to demystify deep learning only the understanding of a basic ANN is needed.

While there are many platforms that allow anybody to leverage the power of A.I. today via such environments as: TensorFlow, PyTorch, Keras, etc., the details of how these systems work is still a mystery to most users. During this presentation we will take a look "under the hood" of the mathematics which is driving all the excitement of the A.I. movement along with the associated challenges. While some knowledge of advanced math concepts would be helpful it is not required as the main math that one needs to be comfortable with is just multiplication and addition. The complexity comes in with the scale and magnitude of the number of operations needed and not the complexity of the operations themselves.

A demonstration will be done with a Python coded example. For those that would like to try it out themselves, the code will be made available during the session.

----

Manuel "Manny" Muro studied Electrical & Computer Engineering at N.C. State University, along with a minor in Business. With more than 25 years of industry experience, he has been involved in just about every aspect of the design, verification and testing of semiconductor devices particularly Microprocessors and SoC/ASIC/FPGAs along with Software, RF, Memory, Analog and Mixed-Signal experiences as well. He has worked at start-up companies, as well as some mid-sized and larger companies. Since 2014, he has been working on the hardware implementation of A.I. systems for a local start-up company in North Plains, Oregon; which resulted in two patents being granted at the end of 2018. More recently he has been doing research and development on coming up with novel hardware approaches to significantly accelerate how machine learning is carried out to address both the speed and power consumption issues associated with current methods in how A.I. systems are trained.

-----

diagram froma; https://cacm.acm.org/magazines/2018/6/228030-deep-learning-hunts-for-signals-among-the-noise/fulltext