CS591 progress bar Storage Layouts NoSQL Engines Rows vs Cols vs - PowerPoint PPT Presentation

CS 591: Da Data S Systems & & M ML Prof. Manos Athanassoulis mathan@bu.edu http://manos.athanassoulis.net/classes/CS591

CS591 progress bar Storage Layouts NoSQL Engines Rows vs Cols vs Hybrid LSM-Trees Distributed DB Hash-based Database Systems New Hardware at Global Scale Indexing Flash Storage Data Skipping MapReduce Multi-core Adaptive Indexing Computing at Scale Indexing Scientific Data Management When to use? In-situ Query Processing UpBit Today: Array Data

CS591 progress bar Data Systems & ML Storage Layouts NoSQL Engines Rows vs Cols vs Hybrid LSM-Trees Distributed DB ML for Systems Hash-based Database Systems Automatic Data New Hardware at Global Scale Indexing System Design Flash Storage Data Skipping MapReduce Multi-core Adaptive Indexing Computing at Scale Indexing Scientific Data Management When to use? In-situ Query Processing Systems for ML UpBit Today: Array Data ML building blocks

CS591 progress bar Data Systems & ML Storage Layouts NoSQL Engines Rows vs Cols vs Hybrid LSM-Trees Distributed DB ML for Systems Hash-based Database Systems Automatic Data New Hardware at Global Scale Indexing System Design Flash Storage Data Skipping MapReduce Multi-core Adaptive Indexing Computing at Scale Indexing Scientific Data Management When to use? In-situ Query Processing Systems for ML UpBit Today: Array Data ML building blocks how can we ef efficien ently s support s statistical queries on large datasets? Data Systems & ML: how can we us analysis of data & queries to be une data systems? use s stat atistical al anal better t r tune

CS591 progress bar CS591 progress bar Data Systems & ML Data Systems & ML Learning Indexes Learning Indexes Storage Layouts Storage Layouts NoSQL Engines NoSQL Engines Rows vs Cols vs Hybrid Rows vs Cols vs Hybrid LSM-Trees LSM-Trees Distributed DB Distributed DB ML for Systems ML for Systems Hash-based Hash-based Database Systems Database Systems Automatic Data Automatic Data New Hardware New Hardware at Global Scale at Global Scale Indexing Indexing System Design System Design Flash Storage Flash Storage Data Skipping Data Skipping Learned Indexes Learned Indexes MapReduce MapReduce Multi-core Multi-core Adaptive Indexing Adaptive Indexing Learn Data Distributions Learn Data Distributions Computing at Scale Computing at Scale Indexing Indexing for Indexing for Indexing Scientific Data Management Scientific Data Management When to use? When to use? Data Calculator Data Calculator In-situ Query Processing In-situ Query Processing Systems for ML Systems for ML UpBit UpBit Synthesize Indexes Synthesize Indexes Today: Array Data Today: Array Data ML building blocks ML building blocks Added video presentations for the last four papers!

Pr Project Pr Presentations April 29 th , 11:59pm: su submi mit p t pro roject re ct report a rt and c code April 30 th and May 2 nd : 6 + 6 10 6 + 6 10-mi minut nute pre prese sentati ations ns (doodle link will be sent after class) May 7 th , 11:59pm (hard deadline): se send u updated re report ( rt (if n needed) (maximum project grade change 10%)

Visitor: Charles Fracchia CEO and Co-Founder, BioBright Inc. Data Management at Scale for Scientific (focus on Biomedical) Discovery Visits MiDAS group on May 3 rd Talk on Friday, May 3 rd , at 11am (room TBA)