Search

We develop variational search distributions (VSD), a method for finding discrete, combinatorial designs of a rare desired class in a batch sequential manner with a fixed experimental budget. We formalize the requirements and desiderata for this problem and formulate a solution via variational inference that fulfill these. In particular, VSD uses off-the-shelf gradient based optimization routines, and can take advantage of scalable predictive models. We show that VSD can outperform existing baseline methods on a set of real sequence-design problems in various biological systems., Comment: 16 pages, 5 figures, Appendix material included

If goals set under the Paris Agreement are met, the world may hold warming well below 2 C; however, parties are not on track to deliver these commitments, increasing focus on policy implementation to close the gap between ambition and action. Recently, the US government passed its most prominent piece of climate legislation to date, the Inflation Reduction Act of 2022 (IRA), designed to invest in a wide range of programs that, among other provisions, incentivize clean energy and carbon management, encourage electrification and efficiency measures, reduce methane emissions, promote domestic supply chains, and address environmental justice concerns. IRA's scope and complexity make modeling important to understand impacts on emissions and energy systems. We leverage results from nine independent, state-of-the-art models to examine potential implications of key IRA provisions, showing economy wide emissions reductions between 43-48% below 2005 by 2035.

Reinforcement learning can learn amortised design policies for designing sequences of experiments. However, current amortised methods rely on estimators of expected information gain (EIG) that require an exponential number of samples on the magnitude of the EIG to achieve an unbiased estimation. We propose the use of an alternative estimator based on the cross-entropy of the joint model distribution and a flexible proposal distribution. This proposal distribution approximates the true posterior of the model parameters given the experimental history and the design policy. Our method overcomes the exponential-sample complexity of previous approaches and provide more accurate estimates of high EIG values. More importantly, it allows learning of superior design policies, and is compatible with continuous and discrete design spaces, non-differentiable likelihoods and even implicit probabilistic models.

Pursuing educational qualifications later in life is an increasingly common phenomenon within OECD countries since technological change and automation continues to drive the evolution of skills needed in many professions. We focus on the causal impacts to economic returns of degrees completed later in life, where motivations and capabilities to acquire additional education may be distinct from education in early years. We find that completing an additional degree leads to more than \$3000 (AUD, 2019) extra income per year compared to those who do not complete additional study. For outcomes, treatment and controls we use the extremely rich and nationally representative longitudinal data from the Household Income and Labour Dynamics Australia survey (HILDA). To take full advantage of the complexity and richness of this data we use a Machine Learning (ML) based methodology for causal effect estimation. We are also able to use ML to discover sources of heterogeneity in the effects of gaining additional qualifications. For example, those younger than 45 years of age when obtaining additional qualifications tend to reap more benefits (as much as \$50 per week more) than others., Comment: 80 pages, 16 figures

Deep learning has been a popular topic and has achieved success in many areas. It has drawn the attention of researchers and machine learning practitioners alike, with developed models deployed to a variety of settings. Along with its achievements, research has shown that deep learning models are vulnerable to adversarial attacks. This finding brought about a new direction in research, whereby algorithms were developed to attack and defend vulnerable networks. Our interest is in understanding how these attacks effect change on the intermediate representations of deep learning models. We present a method for measuring and analyzing the deviations in representations induced by adversarial attacks, progressively across a selected set of layers. Experiments are conducted using an assortment of attack algorithms, on the CIFAR-10 dataset, with plots created to visualize the impact of adversarial attacks across different layers in a network.

Deep learning models have been used for a wide variety of tasks. They are prevalent in computer vision, natural language processing, speech recognition, and other areas. While these models have worked well under many scenarios, it has been shown that they are vulnerable to adversarial attacks. This has led to a proliferation of research into ways that such attacks could be identified and/or defended against. Our goal is to explore the contribution that can be attributed to using multiple underlying models for the purpose of adversarial instance detection. Our paper describes two approaches that incorporate representations from multiple models for detecting adversarial examples. We devise controlled experiments for measuring the detection impact of incrementally utilizing additional models. For many of the scenarios we consider, the results show that performance increases with the number of underlying models used for extracting representations., Comment: Correction: replaced "model-wise" with "unit-wise" in the first sentence of Section 3.2

It has been shown that deep learning models are vulnerable to adversarial attacks. We seek to further understand the consequence of such attacks on the intermediate activations of neural networks. We present an evaluation metric, POP-N, which scores the effectiveness of projecting data to N dimensions under the context of visualizing representations of adversarially perturbed inputs. We conduct experiments on CIFAR-10 to compare the POP-2 score of several dimensionality reduction algorithms across various adversarial attacks. Finally, we utilize the 2D data corresponding to high POP-2 scores to generate example visualizations.

Most work in algorithmic fairness to date has focused on discrete outcomes, such as deciding whether to grant someone a loan or not. In these classification settings, group fairness criteria such as independence, separation and sufficiency can be measured directly by comparing rates of outcomes between subpopulations. Many important problems however require the prediction of a real-valued outcome, such as a risk score or insurance premium. In such regression settings, measuring group fairness criteria is computationally challenging, as it requires estimating information-theoretic divergences between conditional probability density functions. This paper introduces fast approximations of the independence, separation and sufficiency group fairness criteria for regression models from their (conditional) mutual information definitions, and uses such approximations as regularisers to enforce fairness within a regularised risk minimisation framework. Experiments in real-world datasets indicate that in spite of its superior computational efficiency our algorithm still displays state-of-the-art accuracy/fairness tradeoffs., Comment: arXiv admin note: text overlap with arXiv:2001.06089

Algorithmic fairness involves expressing notions such as equity, or reasonable treatment, as quantifiable measures that a machine learning algorithm can optimise. Most work in the literature to date has focused on classification problems where the prediction is categorical, such as accepting or rejecting a loan application. This is in part because classification fairness measures are easily computed by comparing the rates of outcomes, leading to behaviours such as ensuring that the same fraction of eligible men are selected as eligible women. But such measures are computationally difficult to generalise to the continuous regression setting for problems such as pricing, or allocating payments. The difficulty arises from estimating conditional densities (such as the probability density that a system will over-charge by a certain amount). For the regression setting we introduce tractable approximations of the independence, separation and sufficiency criteria by observing that they factorise as ratios of different conditional probabilities of the protected attributes. We introduce and train machine learning classifiers, distinct from the predictor, as a mechanism to estimate these probabilities from the data. This naturally leads to model agnostic, tractable approximations of the criteria, which we explore experimentally., Comment: Accepted to the 2nd Ethics of Data Science Conference 2020 (March, Sydney, Australia)

Spatial data about individuals or businesses is often aggregated over polygonal regions to preserve privacy, provide useful insight and support decision making. Given a particular aggregation of data (say into local government areas), the re-aggregation problem is to estimate how that same data would aggregate over a different set of polygonal regions (say electorates) without having access to the original unit records. Data61 is developing new re-aggregation algorithms that both estimate confidence intervals of their predictions and utilize additional related datasets when available to improve accuracy. The algorithms are an improvement over the current re-aggregation procedure in use by the ABS, which is manually applied by the data user, less accurate in validation experiments and provides a single best guess answer. The algorithms are deployed in an accessible web service that automatically learns a model and applies it to user-data. This report formulates the re-aggregation problem, describes Data61's new algorithms, and presents preliminary validation experiments.