We have two great talks scheduled for June that will explore the application of both neural networks and machine learning in the financial markets. We will have our usual networking session after the talks with food and drink generously provided by our sponsors.
Many of you have requested we videotape these events. For this meetup I’m going to see if we can crowdfund that cost. For those of you who contribute $10 or more we will send a link to the video once it’s taped. I’ve set up an Eventbrite page to collect funds.
Professor Mark Kon presents: “Some background on machine learning and its uses in financial prediction”
Machine learning is still a growing field in both its theoretical aspects and its applications. It is being applied to such areas as robotics, self-driving cars, and computational biology/drug discovery. Its origins lie in artificial neural network theory, which began an evolution in the 90s toward the broader field of machine learning. The area is based on some simple principles which I will discuss. The first is the universality of feature vectors, i.e., the possibility of encoding all objects into strings of numbers. The second is the "geometrization" of learning, i.e., the translation of learning problems into geometrical location tasks, and the third is the so-called kernel trick. I will explain the basic ideas and give examples of applications in financial and other areas.
About the presenter:
Mark Kon is a professor of Mathematics and Statistics at Boston University, and is also a faculty member in the Neuroscience and Bioinformatics programs. His areas of research have included neural networks, machine learning, computational biology, and quantum computation. He studied mathematics, physics, and psychology at Cornell, and mathematics at MIT.
Eric Morris presents “Introduction to Neural Networks in Financial Market Analysis”
Neural networks for machine learning have a rich history and recent resurgence in applied problems across a wide range of fields, especially where it is challenging to write rule-based programs. In this talk we will briefly explore the biological foundation of neural networks and why the human brain can be a highly advantageous paradigm from which to construct analytic hypotheses. After an introduction to biological neural networks, we will explore the application of these characteristics to artificial neural networks and their components. Finally, we will use what we have learned to demonstrate a practical application of applying neural networks to financial data to generate a predictive hypothesis.
About the presenter:
Eric Morris is consulting out of the Cambridge Innovation Center and working on several projects in the finance and technology space. He recently left the French firm Capgemini where he was a consultant on strategy and technology to Fortune 500 clients. Eric graduated with honors from Cornell University where he earned a double major studying Biology (concentrating in biotechnology and business) and Applied Economics and Management in Cornell's College of Agriculture and Life Sciences School and Dyson School of Applied Economics and Management respectively.