On the face of it, it seems like some kind of impossible oxymoron. But it makes perfect sense… (Or at least it does to me now that I've attended the NZ Tech Marketers September event, Unleash your creativity with the power of AI.) Auckland's Tech Marketer contingent had the benefit of insights from Amanda Johnston-Pell, IBM's Chief Marketing Officer for Australia and New Zealand, and our Wellington and Christchurch cohort were joined by the ever-impassioned Isuru (Issy) Fernando, IBM New Zealand's Chief Design and Technology Officer. Both shared findings from the recent IBM 2019 Marketing Trends report: Nine factors reshaping marketing and how you can stay ahead of them. Doing this makes it less scary, binary and wo/man v. machine-ish. Issy says: "There's a lot of hype and uncertainty – and a lot of fud – out there about what AI is. We need to think of AI as less artificial reality and more augmented reality. It really changes the conversation. "If you look at humans across civilisation we've been augmenting ourselves with machines all the time.

SEEK Limited (ASX: SEK) is one of Australia's most entrepreneurial digital businesses and also one of the heaviest investors in new tech, product development, and start up or early stage ventures (ESVs) for long-term growth. In fact it's ready to wear $25 million to $30 million in ESV losses over FY 2020 such is it's commitment to sacrificing the short term for long term success. It now has an aspirational revenue of $5 billion by FY 2025 versus the $1.54 billion delivered in FY 2019, which would be an impressive result if achieved. It recently reported how it's an Australian market leader in artificial intelligence (AI) investment with a team of more than 100 specialist data scientists and software engineers building AI that learns from how candidates search job ads to better target ads for advertisers. According to SEEK the new AI has resulted in an 11% increase in job ad click through rate, a 10% uplift in candidate applications per session and 600,000 more applications per month across its platforms.

We propose $\tt RandUCB$, a bandit strategy that uses theoretically derived confidence intervals similar to upper confidence bound (UCB) algorithms, but akin to Thompson sampling (TS), uses randomization to trade off exploration and exploitation. In the $K$-armed bandit setting, we show that there are infinitely many variants of $\tt RandUCB$, all of which achieve the minimax-optimal $\widetilde{O}(\sqrt{K T})$ regret after $T$ rounds. Moreover, in a specific multi-armed bandit setting, we show that both UCB and TS can be recovered as special cases of $\tt RandUCB.$ For structured bandits, where each arm is associated with a $d$-dimensional feature vector and rewards are distributed according to a linear or generalized linear model, we prove that $\tt RandUCB$ achieves the minimax-optimal $\widetilde{O}(d \sqrt{T})$ regret even in the case of infinite arms. We demonstrate the practical effectiveness of $\tt RandUCB$ with experiments in both the multi-armed and structured bandit settings. Our results illustrate that $\tt RandUCB$ matches the empirical performance of TS while obtaining the theoretically optimal regret bounds of UCB algorithms, thus achieving the best of both worlds.

Conventional machine learning algorithms cannot be applied until a data matrix is available to process. When the data matrix needs to be obtained from a relational database via a feature extraction query, the computation cost can be prohibitive, as the data matrix may be (much) larger than the total input relation size. This paper introduces Rk-means, or relational k -means algorithm, for clustering relational data tuples without having to access the full data matrix. As such, we avoid having to run the expensive feature extraction query and storing its output. Our algorithm leverages the underlying structures in relational data. It involves construction of a small {\it grid coreset} of the data matrix for subsequent cluster construction. This gives a constant approximation for the k -means objective, while having asymptotic runtime improvements over standard approaches of first running the database query and then clustering. Empirical results show orders-of-magnitude speedup, and Rk-means can run faster on the database than even just computing the data matrix.

We consider stochastic second order methods for minimizing strongly-convex functions under an interpolation condition satisfied by over-parameterized models. Under this condition, we show that the regularized sub-sampled Newton method (R-SSN) achieves global linear convergence with an adaptive step size and a constant batch size. By growing the batch size for both the sub-sampled gradient and Hessian, we show that R-SSN can converge at a quadratic rate in a local neighbourhood of the solution. We also show that R-SSN attains local linear convergence for the family of self-concordant functions. Furthermore, we analyse stochastic BFGS algorithms in the interpolation setting and prove their global linear convergence. We empirically evaluate stochastic L-BFGS and a "Hessian-free" implementation of R-SSN for binary classification on synthetic, linearly-separable datasets and consider real medium-size datasets under a kernel mapping. Our experimental results show the fast convergence of these methods both in terms of the number of iterations and wall-clock time.

Estimation of individual treatment effects is often used as the basis for contextual decision making in fields such as healthcare, education, and economics. However, in many real-world applications it is sufficient for the decision maker to have upper and lower bounds on the potential outcomes of decision alternatives, allowing them to evaluate the trade-off between benefit and risk. With this in mind, we develop an algorithm for directly learning upper and lower bounds on the potential outcomes under treatment and non-treatment. Our theoretical analysis highlights a trade-off between the complexity of the learning task and the confidence with which the resulting bounds cover the true potential outcomes; the more confident we wish to be, the more complex the learning task is. We suggest a novel algorithm that maximizes a utility function while maintaining valid potential outcome bounds. We illustrate different properties of our algorithm, and highlight how it can be used to guide decision making using two semi-simulated datasets.

We propose a general method for distributed Bayesian model choice, where each worker has access only to non-overlapping subsets of the data. Our approach approximates the model evidence for the full data set through Monte Carlo sampling from the posterior on every subset generating a model evidence per subset. The model evidences per worker are then consistently combined using a novel approach which corrects for the splitting using summary statistics of the generated samples. This divide-and-conquer approach allows Bayesian model choice in the large data setting, exploiting all available information but limiting communication between workers. Our work thereby complements the work on consensus Monte Carlo (Scott et al., 2016) by explicitly enabling model choice. In addition, we show how the suggested approach can be extended to model choice within a reversible jump setting that explores multiple models within one run.

Contrastive unsupervised representation learning (CURL) is the state-of-the-art technique to learn representations (as a set of features) from unlabelled data. While CURL has collected several empirical successes recently, theoretical understanding of its performance was still missing. In a recent work, Arora et al. ( 2019) provide the first generalisation bounds for CURL, relying on a Rademacher complexity. We extend their framework to the flexible PAC-Bayes setting, allowing to deal with the non-iid setting. We present PAC-Bayesian generalisation bounds for CURL, which are then used to derive a new representation learning algorithm. Numerical experiments on real-life datasets illustrate that our algorithm achieves competitive accuracy, and yields generalisation bounds with non-vacuous values.

Smarter Search is an inside created reason fabricated search motor, which uses AI to adaptively and self-sufficiently gain from applicants' searching hirers' activity advertisements on SEEK. This venture has driven the conveyance of increasingly important search results from each time an applicant searches the stage. These advantages of Smarter Search for applicants and hirers stretch out over SEEK's foundation, regardless of whether got to on portable or work area – which is urgent as more than 71 percent of visits to SEEK happen on a cell phone (31 percent versatile web and 39%). Application appearance has expanded by 25 percent and 16 percent year-on-year, with 52 percent of all applications presently submitted through portable. With 22 years of experience helping hirers and competitors associate through computerized stages, SEEK has utilized this information and family in occupation market understanding, joined with the abilities of their artificial intelligence group, to fabricate an altered AI-based search motor.

Dell Technologies Australia and New Zealand managing director Angela Fox has painted a future where humans and machines learn to live in harmony and machines evolve to be consumers. Delivering the Dell Technologies Forum keynote in Sydney last week, Fox discussed research that was conducted with the Institute of the Future, which looked at the next era of human-machine partnerships. Fox touched on three developments that she expects will shift the economy in the future, with the first being autonomous commerce. "We believe that you'll see machines evolving into consumers. They will use a mix of sensors, software updates, and artificial intelligence (AI) to determine when they -- and the people they serve -- are functioning sub-optimally, but more importantly, they will find ways to remedy it autonomously," Fox said.