Font Size: 

Operational Research (OR) is a discipline that is committed to the design and implementation of advanced analytical methods to support better decisionmaking. Applied OR supports decision-making by promoting the use of the tools, modeling programs, and hands-on experience needed to solve real-life resource allocation problems.  Quantitative finance (quant-fin) is a professional branch within applied OR.  Today, the shift within quant-fin from traditional parametric modeling (OLS regression, Logit methods, ANOVA, etc.) to the nonparametric methods of 21st century algorithms involving machine learning and artificial intelligence (AI) is not always an intuitive one. The purpose of this hands-on workshop is to offer attendees direct access to the real-time implementation of AI for econometric modeling, forecasting and classification. Researchers and educators working in the fields of price prediction (equities, futures, options, etc.), modeling educational assessment, ESG stock classification, and more will find the state-of-the-art results from using new OR-based algorithms intuitive and newly insightful.  Specifically, this hands-on demonstration will feature the use uni- and multivariate radial basis function artificial neural networks (RANNs) and Kohonen self-organizing maps.  The RANNs will be demonstrate with alternative transfer functions (e.g., standardize vs. multiquadric, etc.) that are tailored to generate solutions for mapping, prediction, and discrete choice.  Specifically, at a minimum, the demonstration will use an Excel datafile for input to show the difference between an OLS regression and a RANN regression.  Also planned is a demonstration of how use intelligent algorithms to classify ESG generated stock residual returns. We compare solutions generated by factor analysis to those produced by application of the RANN (softmax transfer function). The concepts demonstrated in the workshop are rich in research ideology but, as the supporting documentation presents, all demonstrated methods are easily incorporated in the senior/graduate-level business classroom (see the appendix for a student submitted homework assignment). Attendees are encouraged to bring an Internet-connected device (preferably a computer or tablet) to the workshop.  Attendees are also encouraged to submit a dataset prior to the start of the conference.  Up to two submitted datasets will be used to demonstrate the methods presented in this workshop (note: workshop presenters are willing to abide by an accompanying NDA).  This 2018 workshop is designed to be a hands-on extension of session 33 on the 2017 NABET conference program, “Active and Experiential Learning in the Evolving Quant-FIN Classroom.â€

Finance, education, ethics, management, supply chain, statistics