Not all neural network architectures are created equal, some perform much better than others for certain tasks. But how important are the weight parameters of a neural network compared to its architecture? In this work, we question to what extent neural network architectures alone, without learning any weight parameters, can encode solutions for a given task. We propose a search method for neural network architectures that can already perform a task without any explicit weight training. To evaluate these networks, we populate the connections with a single shared weight parameter sampled from a uniform random distribution, and measure the expected performance. We demonstrate that our method can find minimal neural network architectures that can perform several reinforcement learning tasks without weight training. On a supervised learning domain, we find network architectures that achieve much higher than chance accuracy on MNIST using random weights.

Smartphones have become the ultimate'personal' computer, yet despite this, general-purpose data mining and knowledge discovery tools for mobile devices are surprisingly rare. DataLearner is a new data mining application designed specifically for Android devices that imports the Weka data mining engine and augments it with algorithms developed by Charles Sturt University. Moreover, DataLearner can be expanded with additional algorithms. Combined, DataLearner delivers 40 classification, clustering and association rule mining algorithms for model training and evaluation without need for cloud computing resources or network connectivity. It provides the same classification accuracy as PCs and laptops, while doing so with acceptable processing speed and consuming negligible battery life. With its ability to provide easy-to-use data mining on a phone-size screen, DataLearner is a new portable, self-contained data mining tool for remote, personalised and educational applications alike. DataLearner features four elements - this paper, the app available on Google Play, the GPL3-licensed source code on GitHub and a short video on YouTube.

A key component of most neural network architectures is the use of normalization layers, such as Batch Normalization. Despite its common use and large utility in optimizing deep architectures that are otherwise intractable, it has been challenging both to generically improve upon Batch Normalization and to understand specific circumstances that lend themselves to other enhancements. In this paper, we identify four improvements to the generic form of Batch Normalization and the circumstances under which they work, yielding performance gains across all batch sizes while requiring no additional computation during training. These contributions include proposing a method for reasoning about the current example in inference normalization statistics which fixes a training vs. inference discrepancy; recognizing and validating the powerful regularization effect of Ghost Batch Normalization for small and medium batch sizes; examining the effect of weight decay regularization on the scaling and shifting parameters γ and β; and identifying a new normalization algorithm for very small batch sizes by combining the strengths of Batch and Group Normalization.

A group of Australian researchers have recently developed an AI-powered smartphone app that can diagnose respiratory disorders by "listening" to the user's cough. This technology was developed by researchers at Curtin University and The University of Queensland, Australia, whose findings were published June 6 in the journal Respiratory Research. The researchers created an algorithm that can analyze coughs for features that are unique to five different diseases. This technique is similar to speech recognition technologies in that the software examines the auditory cough for characteristics specific to these conditions. This is typically done by a physician during a clinical exam, with a stethoscope being used to listen to sound produced while breathing or coughing (auscultation). The downside to this is that the patient must be in the presence of a trained professional to have their respiration sounds analyzed.

Multi-hop Reading Comprehension (RC) requires reasoning and aggregation across several paragraphs. We propose a system for multi-hop RC that decomposes a compositional question into simpler sub-questions that can be answered by off-the-shelf single-hop RC models. Since annotations for such decomposition are expensive, we recast sub-question generation as a span prediction problem and show that our method, trained using only 400 labeled examples, generates sub-questions that are as effective as human-authored sub-questions. We also introduce a new global rescoring approach that considers each decomposition (i.e. the sub-questions and their answers) to select the best final answer, greatly improving overall performance. Our experiments on HotpotQA show that this approach achieves the state-of-the-art results, while providing explainable evidence for its decision making in the form of sub-questions.

Assessing how individuals perform different activities is key information for modeling health states of individuals and populations. Descriptions of activity performance in clinical free text are complex, including syntactic negation and similarities to textual entailment tasks. We explore a variety of methods for the novel task of classifying four types of assertions about activity performance: Able, Unable, Unclear, and None (no information). We find that ensembling an SVM trained with lexical features and a CNN achieves 77.9% macro F1 score on our task, and yields nearly 80% recall on the rare Unclear and Unable samples. Finally, we highlight several challenges in classifying performance assertions, including capturing information about sources of assistance, incorporating syntactic structure and negation scope, and handling new modalities at test time. Our findings establish a strong baseline for this novel task, and identify intriguing areas for further research.

NASCAR will use the breadth and depth of AWS technologies to build cloud-based services and automate processes, including a new video series on NASCAR.com The video series will debut heading into the Monster Energy NASCAR Cup Series race at Michigan International Speedway, sharing the greatest historical moments in NASCAR racing with viewers. NASCAR is migrating its 18-petabyte video archive to AWS, and will leverage Amazon Rekognition--an AWS service that adds intelligent image and video analysis to applications--to automatically tag specific video frames with metadata, such as driver, car, race, lap, time, and sponsors so they can easily search those tags to surface the most iconic moments from past races. By using AWS's services, NASCAR expects to save thousands of hours of manual search time each year, and will be able to easily surface flashbacks like Dale Earnhardt Sr.'s 1987 "Pass in the Grass" or Denny Hamlin's 2016 Daytona 500 photo finish, and quickly deliver these to fans via video clips on NASCAR.com and social media channels. NASCAR will leverage AWS services to enhance its full range of media assets including websites, mobile applications, and social properties for its 80 million fans worldwide.

Agerris, an Australia-based robotics and AI platform for agriculture, announced over the weekend that it has raised $6.5 million (AUSD) in seed funding from Uniseed, Carthona Capital and BridgeLane Group. The startup was founded by Professor Salah Sukarrieh and began as research at the Australian Centre for Field Robotics at the University of Sydney (which is also a partner in Uniseed). From the looks of it, Agerris is building a modular robotics and AI platform that has broad applications for both plant and livestock farmers. According to a University of Sydney news post, Agerris has two main products. The "Swagbot" can autonomously monitor and identify weed issues, detect food and crops through computer vision, as well as herd livestock.

Microsoft Australia's national technology officer Lee Hickin has reaffirmed that artificial intelligence (AI) technologies are not replacing humans in the workforce, but are placing humans into positions where they can provide more value in their work. Speaking at ZDNet's Next Big Thing event in Sydney on Thursday, Hickin said the implementation of AI into work environments will augment the human experience, rather than replace it altogether. Using Microsoft's work at Northern Territory fisheries as an example, Hickin said the implementation of AI can "take away what we would call'grunt work' in jobs and functions". The AI fisheries project uses the company's Azure Cognitive Service to identify and count fish in waters without needing to sort through hours of under-water footage. The solution has already shown that the local golden snapper and black jewfish species are overfished.

Tinder is giving users more tools to express their sexuality. The dating app announced on Tuesday that users can now select up to three terms that they most identify with from a list of nine options. Tinder is giving users more tools to express their sexuality. Users can choose from nine orientations, including straight, gay, lesbian, bisexual, asexual, demisexual, pansexual, queer and questioning. From there, they can decide whether they want that information to show up on their public-facing profile.