12 August 2016
Data Mining for Data Science
Data science is an emerging interdisciplinary area that studies how to develop and use methods for extracting useful information and knowledge from data. Patterns and models mined from various kinds of data can be descriptive, predictive or explanatory. This course offers an introduction to data mining approaches for knowledge discovery through accessible lectures and hands-on practice assignments to be completed in small groups.
The introduction to the field will exemplify the goals (to explore, to explain, to identify/detect or to predict), the means (classification/ prediction, clustering, pattern mining, anomaly and outlier detection, etc.) and the ways to assess the outputs of the knowledge discovery through data mining .
The overview of the key state-of-the-art data mining techniques will be focused on predictive analytics. However, we will discuss also how to compute similarities, to perform clustering and pattern mining on data of different nature (e.g. text, transactions, event logs, sensory measurements, complex networks) and how to use that in data-driven decision support or decision making.
Several techniques that are essential for applying data mining in practice, particularly in the context of data science, will be taught through illustrative cases studies. We will discussed the main ideas for and challenges in persuasion and recommendation, cost-sensitive learning, uplift modeling, learning from A/B testing, responsible (privacy- and ethics –aware, trustworthy and accountable) data mining.
The closing lecture will be devoted to the revision of the opportunities that data mining provides for data science, the pitfalls that must be avoided, and the trade-offs that are inherently hard to avoid.
Lecturer: Associate Professor Mykola Pechenizkiy (TU Eindhoven, the Netherlands)
Coordinators: Prof. Tommi Kärkkäinen and Dr. Sanna Mönkölä (University of Jyvaskyla)
The Summer School annually offers courses for advanced master’s students, graduate students, and post-docs in the various fields of science and information technology.
Prerequisites: Basic knowledge of matrix algebra and probability theory would be helpful, but not strictly required. Basic programming skills (in some language like Python, R or Matlab) are essential for completing the practical assignments. It would be easier for you to follow the course if you already took an introduction to data mining. However, this course is self-contained.
The most important aims of the Summer School are to develop post-graduates scientific readiness and to offer students the possibility to study in a modern, scientific environment and to create connections to the international science community. The Summer School offers an excellent pathway to develop international collaboration in post-graduate research.
Passing: Obligatory attendance at lectures, and completing the exercises.
Grading: Pass/fail; 60% or more of the full grade for passing.
Group work: 60% (3 assignments, 20 % each):
EUR 0: : Participating the Summer School is free of charge, but student have to cover the costs of own travel, accommodation and meals at Jyväskylä.
The 26th Jyväskylä Summer School is not able to grant any Summer School students financial support.