optimization of integration flows  Systems support for time series forecasting  In-memory indexing and query processing Data Management Group DB group https://github.com/ apache/systemds 6 706.015 Introduction Large-Scale Analysis (DIA, WS) Master Bachelor Data management from user/application perspective Distributed Data Management ML system internals DB system internals + prog. project Prog. projects converge to an optimal model. It is crucial for performance to fit the data into single-node or distributed main memory. % 2. Say why it's an interesting problem General-purpose, heavy- and lightweight

TPC-DS, JOB, MLPerf  #4 End-to-end Applications  Evaluate in larger scope of real datasets and query workloads  Examples: Customer workload, ML pipelines (dataprep, training, eval) Experiments and agreed benchmarks  Benchmarks can define a field and foster innovation  #1 Data Management  Query processing: 007, TPC-C, TPC-E, TPC-H, TPC-DS (w/ audit)  Join ordering: JOB  #2 “Big Data” WS 2021/22 Presentation – Experimental Setting  Hardware Selection  Multiple nodes for distributed computation  Avoid too outdated HW (irrelevance)  Find Balanced Level of Detail  Underspecified:

Type Information (RTTI) Avoid using Run Time Type Information (RTTI). RTTI allows a programmer to query the C++ class of an object at run time. This is done by use of typeid or dynamic_cast. Querying the hpp) and parallel/distributed algorithms and data structures (boost/graph/parallel/* and boost/graph/distributed/*). • Property Map from boost/property_map, except parallel/distributed property maps (b

only allowed between assignments. GOOD: PredictCTR <- function(query, property, num.days, show.plot = TRUE) BAD: PredictCTR <- function(query, property, num.days, show.plot = TRUE) Ideally, unit tests

is introduced. Later, Raman and Swart investigated query processing on heavyweight Huffman coding schemes [40], Later, [40] investigated query processing on heavyweight Huffman coding schemes, 22