Adaptive Business Intelligence (ABI): Presentation of the Unit

Transcription

1 Adaptive Business Intelligence (ABI): Presentation of the Unit MAP-i PhD (Edition 2015/16) Lecture Team: Manuel Filipe Santos (University of Minho); Paulo Cortez (University of Minho); Rui Camacho (University of Porto);

2 > Motivation... The Age of Big Data! (and Data Science, Analytics, Data Mining, Business Intelligence, Decision Support Systems,...)

3 The Age of Data Science: Why your kids will want to be data scientists (CNBC News, 3 Jun 2014). Data Scientist is identified as the "sexiest job of the 21st century (Harvard Business Review, Oct 2012). Trends That Will Impact IT in 2015: Analytics for Evidencebased Decision-Making (

4 > Business Intelligence (BI) Business Intelligence (BI) is an umbrella term that includes methodologies, architectures, tools, applications and technologies to enhance decision making. The goal of BI is to: access data from multiple sources, transform these data into information, knowledge and actions. Current trends: real-time access to data (get the right data at the right time and place), use of Web data, use of more Intelligence,...

5 > Adaptive Business Intelligence (ABI) Useful for complex business problems, such as: distribution of cars in USA, marketing campaigns, investment strategies, credit card fraud, online news enhancement,...

6 > Program Introductory ABI concepts: characteristics of complex business problems, BI and ABI, ABI case studies, data mining, prediction, optimization and adaptability. Modern forecasting and optimization methods for ABI: data mining, supervised learning (e.g., neural networks, support vector machine), clustering (e.g. hierarchical and relational clustering), inductive logic programming, heuristic search (e.g. hill-climbing, tabusearch, evolutionary computation), univariate and multivariate time series forecasting. Exploration of ABI tools (e.g., R, WEKA, Evolution Machine, SCS- C) when applied to real-world problems (e.g., Finance, Economy, Marketing, Science, Engineering). ABI can be an interesting complement for KDD (Knowledge Discovery from Databases)

7 > Teaching Methodology and Evaluation Four teaching methodologies will be applied: 1 Lecture exposition of key ABI issues. 2 Active learning (e.g., think-pair-share, in-class teams). 3 Case-based learning. 4 Project based learning; Evaluation will include two elements: A - review of an advanced ABI research article from an ISI journal, leading into a presentation and short article (30%); B - an ABI project (group of 2/3 students) that describes the application of ABI tools to a real-world dataset (70%);

8 > Past ABI grades: Minimum grade of 15 valores (75%; assumes that students performed both the review and project assigments). Maximum grade of 19 valores (95%)

9 > Student Feedback Average of the all past ABI editions (anonymous questionnaires):

10 > Student Feedback Qualitative feedback from the 2012/13 ABI edition: I really enjoyed participating in Advance Business Intelligence class in previous semester, and that is really usable and useful for me, insetting point was interaction between students and teacher and also friendly atmosphere was very nice. > Previous ABI Editions 2014/15: 13 students (University of Porto) 2013/14: 10 students (University of Minho); 2012/13: 8 students (University of Aveiro); 2011/12: 10 students (University of Porto); 2010/11: 6 students (University of Minho); 2007/08: 7 students (University of Porto);

11 > 2014/15 ABI Projects: Stock Market Prediction and Optimization Predicting Popularity in Online News Decide whether a customer should be given or refused credit Community Crime Prediction and Resources Distribution Optimization > Two of these projects were later published as full articles in: 17th Portuguese Conference on Artificial Intelligence (EPIA 2015), Springer LNCS (indexed at Scopus, ISI)

12 > Lecture Team: Manuel Filipe Santos, University of Minho ( Paulo Cortez, University of Minho ( Rui Camacho, University of Porto (

13 > Bibliography Michalewicz, Z., Schmidt, M., Michalewicz, M., and Chiriac, C. (2006). Adaptive Business Intelligence. Springer. P. Cortez (2014). Modern Optimization with R. Springer.