Protocols and Structures for Inference: A RESTful API for Machine Learning

James Montgomery, Mark D. Reid, Barry Drake
Proceedings of The 2nd International Conference on Predictive APIs and Apps, PMLR 50:29-42, 2016.

Abstract

Diversity in machine learning APIs (in both software toolkits and web services), works against realising machine learning’s full potential, making it difficult to draw on individual algorithms from different products or to compose multiple algorithms to solve complex tasks. This paper introduces the Protocols and Structures for Inference (PSI) service architecture and specification, which presents inferential entities—relations, attributes, learners and predictors—as RESTful web resources that are accessible via a common but flexible and extensible interface. Resources describe the data they ingest or emit using a variant of the JSON schema language, and the API has mechanisms to support non-JSON data and future extension of service features.

Cite this Paper

BibTeX
@InProceedings{pmlr-v50-montgomery15, title = {Protocols and Structures for Inference: A RESTful API for Machine Learning}, author = {Montgomery, James and Reid, Mark D. and Drake, Barry}, booktitle = {Proceedings of The 2nd International Conference on Predictive APIs and Apps}, pages = {29--42}, year = {2016}, editor = {Dorard, Louis and Reid, Mark D. and Martin, Francisco J.}, volume = {50}, series = {Proceedings of Machine Learning Research}, address = {Sydney, Australia}, month = {06--07 Aug}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v50/montgomery15.pdf}, url = {https://proceedings.mlr.press/v50/montgomery15.html}, abstract = {Diversity in machine learning APIs (in both software toolkits and web services), works against realising machine learning’s full potential, making it difficult to draw on individual algorithms from different products or to compose multiple algorithms to solve complex tasks. This paper introduces the Protocols and Structures for Inference (PSI) service architecture and specification, which presents inferential entities—relations, attributes, learners and predictors—as RESTful web resources that are accessible via a common but flexible and extensible interface. Resources describe the data they ingest or emit using a variant of the JSON schema language, and the API has mechanisms to support non-JSON data and future extension of service features.} }
Endnote
%0 Conference Paper %T Protocols and Structures for Inference: A RESTful API for Machine Learning %A James Montgomery %A Mark D. Reid %A Barry Drake %B Proceedings of The 2nd International Conference on Predictive APIs and Apps %C Proceedings of Machine Learning Research %D 2016 %E Louis Dorard %E Mark D. Reid %E Francisco J. Martin %F pmlr-v50-montgomery15 %I PMLR %P 29--42 %U https://proceedings.mlr.press/v50/montgomery15.html %V 50 %X Diversity in machine learning APIs (in both software toolkits and web services), works against realising machine learning’s full potential, making it difficult to draw on individual algorithms from different products or to compose multiple algorithms to solve complex tasks. This paper introduces the Protocols and Structures for Inference (PSI) service architecture and specification, which presents inferential entities—relations, attributes, learners and predictors—as RESTful web resources that are accessible via a common but flexible and extensible interface. Resources describe the data they ingest or emit using a variant of the JSON schema language, and the API has mechanisms to support non-JSON data and future extension of service features.
RIS
TY - CPAPER TI - Protocols and Structures for Inference: A RESTful API for Machine Learning AU - James Montgomery AU - Mark D. Reid AU - Barry Drake BT - Proceedings of The 2nd International Conference on Predictive APIs and Apps DA - 2016/06/05 ED - Louis Dorard ED - Mark D. Reid ED - Francisco J. Martin ID - pmlr-v50-montgomery15 PB - PMLR DP - Proceedings of Machine Learning Research VL - 50 SP - 29 EP - 42 L1 - http://proceedings.mlr.press/v50/montgomery15.pdf UR - https://proceedings.mlr.press/v50/montgomery15.html AB - Diversity in machine learning APIs (in both software toolkits and web services), works against realising machine learning’s full potential, making it difficult to draw on individual algorithms from different products or to compose multiple algorithms to solve complex tasks. This paper introduces the Protocols and Structures for Inference (PSI) service architecture and specification, which presents inferential entities—relations, attributes, learners and predictors—as RESTful web resources that are accessible via a common but flexible and extensible interface. Resources describe the data they ingest or emit using a variant of the JSON schema language, and the API has mechanisms to support non-JSON data and future extension of service features. ER -
APA
Montgomery, J., Reid, M.D. & Drake, B.. (2016). Protocols and Structures for Inference: A RESTful API for Machine Learning. Proceedings of The 2nd International Conference on Predictive APIs and Apps, in Proceedings of Machine Learning Research 50:29-42 Available from https://proceedings.mlr.press/v50/montgomery15.html.

Related Material