Overview

Description

CALENDAR DESCRIPTION:

Analytical approaches examining user interaction data to understand how interactive systems meet their users' goals are covered. The data preprocessing techniques, feature engineering for summative and temporal characteristics, statistical, data mining and machine learning techniques used to derive insights are compared, focusing on their benefits and pitfalls.

COURSE DETAILS:

The course topics cover the whole data science processing pipeline, presented in the context of interactive systems. The approaches and techniques learned are widely applicable in any data science project. You will get a sense of what the data science field is, what kind of questions are asked, with what goals, and how data science projects are structured. You will understand the goals of the data cleaning step, perform data cleaning, and strategize for dealing with missing data (e.g., estimating duration when the end-point is missing). We will use data visualizations to understand data and present the results effectively. Next, you will practice feature engineering to support the analysis, focusing particularly on time-related features: time windows, counts, sequences, and patterns. You will learn how to build and interpret models using various prediction and classification methods, including linear and logistic regression, clustering, decision trees, random forests and support vector machines (SVG). You will apply what you learn in the practically oriented project; in the process, you will get familiar with widely used Python libraries for data analysis.

COURSE-LEVEL EDUCATIONAL GOALS:

By the end of this course, students will be able to:

• carry out the data analytics process for human-centred systems from beginning to end
• use proper terminology in the field
• understand various types of data and the issues in analyzing each type of data
• identify the techniques used for each step and when a technique is appropriate
• master steps of data cleaning, feature engineering, analytical technique selection, and analysis results interpretation
• apply linear and logistic regressions, k-means and hierarchical clustering, and machine learning models such as decision trees, random forest, and support vector machines (SVM)
• integrate data capture mechanisms into the system design based on the analysis of needs
• use the tools available in the Python ecosystem to carry out the analysis

Grading

Materials

REQUIRED READING:

REQUIRED READING NOTES: