Transcription of Declarative Machine Learning A Classification of Basic ...
{{id}} {{{paragraph}}}
Example: quiz answers
Declarative Machine Learning A Classification of Basic ...
Declarative Machine Learning . A Classification of Basic Properties and Types Matthias Boehm, Alexandre V. Evfimievski, Niketan Pansare, Berthold Reinwald IBM Research Almaden; San Jose, CA, USA. [ ] 19 May 2016. Table 1: Delineation of Types of Declarative ML. ABSTRACT Declarative ML Tasks , MLbase [27, 33], Declarative Machine Learning (ML) aims at the high-level (fixed task) Columbus [42], DeepDive [32]. specification of ML tasks or algorithms, and automatic gen- Declarative ML Algorithms , OptiML [35], SciDB [9, 34]. eration of optimized execution plans from these specifica- (fixed algorithm) SystemML [7, 22], SimSQL [11]. tions. The fundamental goal is to simplify the usage and/or Large-Scale ML Libraries , MLlib [30], Mahout [37], development of ML algorithms, which is especially impor- (fixed plan) MADlib [12, 23], ORE, Rev R. tant in the context of large-scale computations. However, ML systems at different abstraction levels have emerged over Declarative ML: Declarative ML aims at a high-level time and accordingly there has been a controversy about the specification of ML tasks or algorithms to simplify the usage meaning of this general definition of Declarative ML.
This meta information might be built into the system or annotated in case of extensible systems. Overall operation semantics allow to reason about equivalences, alternative ex-ecution strategies, and costs of these alternatives. Property 5. Implementation-Agnostic Operations: The speci cation of ML tasks or algorithms is independent of
Tags:
Information
Domain:
Source:
Link to this page:
Please notify us if you found a problem with this document: