Data-intensive Systems and Algorithms (DAT535)

Content

The emergence of Big Data and Data-intensive Systems as specialized fields in computing has been motivating development of new techniques and technologies needed to extract knowledge from large datasets. Since Hadoop was conceived in 2005, popular interest in data-intensive systems began to grow. It resulted - over time - in a collection of technologies, methodologies, and practices to cover the complete data lifecycle.

This course is a first step to a variety of roles related to data-intensive systems. The core tasks in these roles that we will address are: roles in a data team, low-level algorithm design and implementation (direct implementation of MapReduce jobs), high-level algorithm design and implementation (utilizing one of data processing frameworks e.g. SparkSQL, MLlib), dataflow design (data pipelines), algorithm optimisation, advocating technology application both in technical and non-technical setting, providing introductory training to coworkers.

Learning outcome

Knowledge

• Characterize Hadoop architecture incl. job tracker, task tracker, scheduling issues, communications, and resource management, etc.
• Characterize Spark/Databricks architecture incl. context, cluster manager, worker node, executor, etc.
• Describe elements of Hadoop/Spark ecosystem and identify their applicability
• Describe and compare RDBMS, NOSQL databases, data warehouse, unstructured big data, and keyed files, and show how to apply them to typical data processing problems

Skills

• Assume various roles in a data team
• Use and reconfigure a data processing setup (based on Hadoop/Spark/DataBricks, OpenStack, or other Cloud setup)
• Analyze real-life problems and propose suitable solutions
• Construct and optimize algorithms and dataflows based on relevant tools for typical problems

General qualifications:

• Evaluate, communicate and defend a data-intensive solution w.r.t. relevant criteria

Required prerequisite knowledge

Recommended prerequisites

Bash programming

Administration of Cloud and container-based environments

Databases, SQL

Exam

Coursework requirements

Three assignments

Students start with 3 mandatory assignments that contain programing and system administration. Assignments are to be completed individually. All mandatory assignments must be passed within deadline so that the student has the right to start with the project. The obligatory assignments give access to the project only in the current semester.

Completion of mandatory lab assignments is to be made at the times and in the groups that are assigned and published. Absence due to illness or for other reasons must be communicated as soon as possible to the laboratory personnel. One cannot expect that provisions for completion of the lab assignments at other times are made unless prior arrangements with the laboratory personnel have been agreed upon.

All group members must participate in the project presentation.

Course teacher(s)

Course coordinator:

Laboratory Engineer:

Head of Department:

Method of work

Open for

Course assessment

Literature