Optimization Methods for Big Data Analytics
The course presents some of the latest trends in Big Data Analytics, with a focus on data mining tasks. We will study algorithms that allow us to find regular patterns in datasets, and extract useful and interesting knowledge via descriptive, predictive, and prescriptive data analytics. We will study both supervised (classification, regression) and unsupervised (clustering) learning. We will present and implement state-of-the-art algorithms for these tasks, eventually including metaheuristic algorithms. The course has an hands-on approach, which emphasizes working on real data and going through all the steps of knowledge discovery, including data preparation, data mining, validation of the extracted knowledge, visualization of the results.
After the course, the students are able to
- understand what is knowledge discovery, what is data mining and what type of insights we can algorithmically extract from massive datasets,
- have a clear overview of data mining tasks, and types of learning,
- understand, implement, and adapt state-of-the-art algorithms to perform the most common data mining tasks,
- prepare and clean the data before running it through a data mining algorithm,
- assess and validate the knowledge they extract from the dataset,
- present and visualize this knowledge in an interesting and meaningful way,
- work independently on all phases of a data analytics task.