Democratising Big Machine Learning

Grant number: DP150103710 | Funding period: 2015 - 2018

Completed

Abstract

Technological advances such as cloud computing have disrupted thousands of businesses managing volatile compute loads. While elements of Big Data are now everywhere, still absent are wide-spread solutions for learning from data at scale-Big Machine Learning, the ultimate goal of Big Data. The greatest problems come not from a lack of distributed machine learning algorithms, but rather from preparing the data needed for fitting, evaluating and applying statistical models; often a manual, messy and costly process. This project proposes to develop advanced databases and statistical techniques for scalable and efficient data preparation, with the goal of bringing Big Machine Learning to a much b..

View full description

Related publications (13)

Scholarly works icon

d-blink: Distributed End-to-End Bayesian Entity Resolution

Neil G Marchant, Andee Kaplan, Daniel N Elazar, Benjamin IP Rubinstein, Rebecca C Steorts

2021-02-16

Entity resolution (ER; also known as record linkage or de-duplication) is the process of merging noisy databases, often in the abs..

Scholarly works icon

Needle in a Haystack: Label-Efficient Evaluation under Extreme Class Imbalance

Neil G Marchant, Benjamin IP Rubinstein

2021-01-01

Important tasks like record linkage and extreme classification demonstrate extreme class imbalance, with 1 minority instance to ev..

Scholarly works icon

Sampling Without Compromising Accuracy in Adaptive Data Analysis

Benjamin Fish, Lev Reyzin, Benjamin Rubinstein

2020-01-01

In this work, we study how to use sampling to speed up mechanisms for answering adaptive queries into datasets without reducing th..

Scholarly works icon

Truth inference at scale: A Bayesian model for adjudicating highly redundant crowd annotations

Y Li, BIP Rubinstein, T Cohn

2019-05-13

Crowd-sourcing is a cheap and popular means of creating training and evaluation datasets for machine learning, however it poses th..

Scholarly works icon

Exploiting Worker Correlation for Label Aggregation in Crowdsourcing

Yuan LI, Benjamin Rubinstein, Trevor Cohn

2019-01-01

Crowdsourcing has emerged as a core component of data science pipelines. From collected noisy worker labels, aggregation models th..

Scholarly works icon

Fast manifold landmarking using locality-sensitive hashing

ZMM Aye, BIP Rubinstein, K Ramamohanarao

2018-01-01

© Springer International Publishing AG, part of Springer Nature 2018. Manifold landmarks can approximately represent the low-dimen..

Scholarly works icon

In Search of an Entity Resolution OASIS: Optimal Asymptotic Sequential Importance Sampling

Neil G Marchant, Benjamin IP Rubinstein

2017-08-01

Entity resolution (ER) presents unique challenges for evaluation methodology. While crowdsourcing platforms acquire ground truth, ..

Scholarly works icon

E-Storm: Replication-based State Management in Distributed Stream Processing Systems

Xunyun Liu, Aaron Harwood, Shanika Karunasekera, Benjamin Rubinstein, Rajkumar Buyya

2017-01-01

Apache Storm is a fault-tolerant, distributed inmemory computation system for processing large volumes of high-velocity data in re..

Scholarly works icon

Fast Trajectory Clustering using Hashing Methods

Ivan Sanchez, Zay Maung Maung Aye, Benjamin IP Rubinstein, Kotagiri Ramamohanarao

2016-01-01

There has been an explosion in the usage of trajectory data. Clustering is one of the simplest and most powerful approaches for kn..

Scholarly works icon

Large Scale Metric Learning

Z Aye, R Kotagiri, B RUBINSTEIN

2016-01-01

Many machine learning and pattern recognition algorithms rely heavily on good distance metrics to achieve competitive performance...

Scholarly works icon

Principled Graph Matching Algorithms for Integrating Multiple Data Sources

Duo Zhang, Benjamin IP Rubinstein, Jim Gemmell

2015-10-01

This paper explores combinatorial optimization for problems of max-weight graph matching on multi-partite graphs, which arise in i..

Scholarly works icon

Sub-Merge: Diving Down to the Attribute-Value Level in Statistical Schema Matching

Z Lim, B RUBINSTEIN

2015-01-25

Matching and merging data from conflicting sources is the bread and butter of data integration, which drives search verticals, e-c..

Scholarly works icon

Attribute-Value-Level Matching

Z Lim, B RUBINSTEIN

2014-11-17

We explore normalisation of attribute values across multiple data sources, where attributes could be categorical, numerical, other..

University of Melbourne Researchers