This post is an introduction to time-series forecasting with torch. Central topics are data input, and practical usage of RNNs (GRUs/LSTMs). Upcoming posts will build on this, and introduce increasingly involved architectures.

The course enables novice practitioners who are interested in entering the field of AI to become job-ready engineers in just six months. There will be a skill assessment to evaluate the student's current skills and experience, but the only prerequisites are a basic understanding of mathematics and statistics and, or alternatively, relevant experience in the industry, such as an internship. According to the 2020 IT Skills and Salary Report, 'AI, cognitive computing and machine learning' was reported as the second weakest IT skillset in organisations across the world, presenting serious hiring challenges for IT managers.[1] The lack of skilled AI professionals forms part of a broader technology skills shortage in Australia, with new research commissioned by AWS revealing that Australia will require an additional 6.5 million newly-skilled and reskilled digital workers by 2025 to meet future demand for technology skills (79 per cent more than Australia currently has). Jon Lang, CEO of AIICT said, "We're seeing a severe shortage of skilled AI professionals and in many instances, organisations are simply unable to fill these roles. With AI and ML becoming a critical part of the digital transformation process for many organisations, the demand is on the rise for engineers with these specific skills across a wide range of industries."

The America's Cup, the oldest trophy in international sport, is competitive sailing's most coveted prize. When the 36th edition begins in early March, the race's defending champion, Emirates Team New Zealand, will hit the water having utilized a new crewmember: an AI bot created by McKinsey. Winning the America's Cup has always been as much about technology and innovation as it is about sailing. Boat designs are governed by the competition's "Class Rule, " which leaves small opportunities for design tweaks that might give one team an advantage over another. Those opportunities, if brilliantly seized, can translate into a shining race-day performance.

SCA has become an early-stage investor in Melbourne-based Sonnant Pty Ltd, a transformational artificial intelligence (AI) and machine learning (ML) company that provides content discovery for the spoken word. According to the GfK Australian Share of Audio 2019, consumption of digital audio is growing in Australia and is expected to reach 80% of the population by 2024 and SCA's recent launch of LiSTNR is designed to be a best in class consumer digital audio experience. SCA's in-house technical capability will be enhanced by Sonnant's AI intelligence, providing an in-depth of understanding of SCA content that allows for rich meta data and exciting new user experiences. Sonnant allows SCA to enhance the key product pillars of LiSTNR delivering a personalised experience to each user based on the topics they most enjoy listening to. SCA Head of Digital and Innovation, Chris Johnson, says "SCA is a proud supporter of the Australian entrepreneurial ecosystem, so when we found a company of Sonnant's calibre in Melbourne, we knew that an investment partnership would provide significant long term value to both parties. "We were delighted when Sonnant's tech outperformed several'off the shelf' products we pitted it against in testing.

Visual Place Recognition (VPR) is often characterized as being able to recognize the same place despite significant changes in appearance and viewpoint. VPR is a key component of Spatial Artificial Intelligence, enabling robotic platforms and intelligent augmentation platforms such as augmented reality devices to perceive and understand the physical world. In this paper, we observe that there are three "drivers" that impose requirements on spatially intelligent agents and thus VPR systems: 1) the particular agent including its sensors and computational resources, 2) the operating environment of this agent, and 3) the specific task that the artificial agent carries out. In this paper, we characterize and survey key works in the VPR area considering those drivers, including their place representation and place matching choices. We also provide a new definition of VPR based on the visual overlap -- akin to spatial view cells in the brain -- that enables us to find similarities and differences to other research areas in the robotics and computer vision fields. We identify numerous open challenges and suggest areas that require more in-depth attention in future works.

Several pieces of work have uncovered performance disparities by conducting "disaggregated evaluations" of AI systems. We build on these efforts by focusing on the choices that must be made when designing a disaggregated evaluation, as well as some of the key considerations that underlie these design choices and the tradeoffs between these considerations. We argue that a deeper understanding of the choices, considerations, and tradeoffs involved in designing disaggregated evaluations will better enable researchers, practitioners, and the public to understand the ways in which AI systems may be underperforming for particular groups of people.

A fingerprint scanner is one of the advanced AI tools used almost everywhere schools, offices, banks, airports where ever we work or have some of our data saved. It is the most forwarding step to link everything that relates to the user and to keep it safe it is locked or secured by the fingerprint. We use smartphones or laptops that are secured by fingerprints that cannot be accessed by others except the user. There is a various tool which keeps the data of the fingerprint through which the scanned pattern is matched and the user get the access of it. It is widely used all over the world not just for securing the data, there is a various system which is a link with the fingerprint scanner. Fingerprint attendance system is used widely in schools, offices to keep the record of the available employee, teacher, and students.

Imagine if artificial intelligence (AI) could pick the next Jeff Bezos, Richard Branson, Bill Gates or Elon Musk? It may be just around the corner according to a QUT researcher into entrepreneurship. A QUT study is looking for thousands of people to complete a short online survey asking them to look at photographs and identify which people are entrepreneurs. The results will be compared to the performance of AI models given the same challenge. Led by Professor Martin Obschonka Director of QUT's Australian Centre for Entrepreneurship Research, the aim of the project is to gauge the capacity of modern AI in recognising entrepreneurship potential in a variety of people.

A novel reinforcement learning algorithm is introduced for multiarmed restless bandits with average reward, using the paradigms of Q-learning and Whittle index. Specifically, we leverage the structure of the Whittle index policy to reduce the search space of Q-learning, resulting in major computational gains. Rigorous convergence analysis is provided, supported by numerical experiments. The numerical experiments show excellent empirical performance of the proposed scheme.

Training machine learning (ML) models typically involves expensive iterative optimization. Once the model's final parameters are released, there is currently no mechanism for the entity which trained the model to prove that these parameters were indeed the result of this optimization procedure. Such a mechanism would support security of ML applications in several ways. For instance, it would simplify ownership resolution when multiple parties contest ownership of a specific model. It would also facilitate the distributed training across untrusted workers where Byzantine workers might otherwise mount a denial-of-service by returning incorrect model updates. In this paper, we remediate this problem by introducing the concept of proof-of-learning in ML. Inspired by research on both proof-of-work and verified computations, we observe how a seminal training algorithm, stochastic gradient descent, accumulates secret information due to its stochasticity. This produces a natural construction for a proof-of-learning which demonstrates that a party has expended the compute require to obtain a set of model parameters correctly. In particular, our analyses and experiments show that an adversary seeking to illegitimately manufacture a proof-of-learning needs to perform *at least* as much work than is needed for gradient descent itself. We also instantiate a concrete proof-of-learning mechanism in both of the scenarios described above. In model ownership resolution, it protects the intellectual property of models released publicly. In distributed training, it preserves availability of the training procedure. Our empirical evaluation validates that our proof-of-learning mechanism is robust to variance induced by the hardware (ML accelerators) and software stacks.