- The Humans of Ai Podcast
We are extremely proud to announce the release of our Podcast, “The Humans of Ai” - where we are going to be focussing on the humans behind the algorithms. In our inaugural episode, we talk to Prof. Claude Sammut about how he got started in the field of Robotics, dive down into what inductive logic programming is, discuss the difference between Behavioural Cloning & Imitation Learning, & talk about the International Robocup to be held at the International Convention Centre in Sydney, 2-8th of July 2019. The Humans of Ai podcast is available on iTunes here: https://podcasts.apple.com/au/podcast/the-humans-of-ai/id1464995550 On Spotify here: https://open.spotify.com/show/2RY5mcNl0iAs8HUTNbwT0J On Sticher here: https://www.stitcher.com/s?fid=414486&refid=stpr Or directly off our site here: https://www.thehumansofai.com/2019/05/22/Prof-Claude-Sammut-on-Symbolic-Ai-&-the-International-Robocup.html If you like the content, please feel free to like click & subscribe!
- DSAi study: Bayesian Inference - Statistical Rethinking Study Group
This is a x-post from the DSAi event. For more information head on over to their site here: https://dsai.org.au/courses/ ~ What is Bayesian Inference? Bayesian Statistics is getting quite popular today, but Bayesian Inference is the way scientists did statistics before the 20th century. All our intuition about things like p-values and confidence intervals actually matches the Bayesian equivalents and not the actual definitions. However, the chance to study Bayesian Inference is quite fleeting and so there are many misconceptions and a general lack of knowledge on the how and the advantages inherent to such an approach. Why bother? DSAi & Varun are running a study group, going through Richard McElreath’s fantastic course, “Statistical Rethinking” in which build up an alternate Bayesian view of data, statistics and inference and question our existing assumptions and standard procedures. This study group is perfect for all levels of knowledge, senior practitioners will have the ability to develop their fundamentals and add to their toolbox, while juniors with fewer assumptions will benefit from the paradigm shift of thinking required to be a Bayesian. In part 1 of 3 we aim to cover: -The relationship between data, models and hypotheses What priors, likelihoods and posteriors are. -Using R and the rethinking package to do simple inference -Building Bayesian equivalents of frequentist models -Understanding and handling causation in models Overfitting, model assessment and motivations of each. Where & when? Date might change. We will meet once a week 6:00m- 8:00 pm for 5 weeks Cost? The event is free to attend with limited seats. Who are we? DSAi is Non-for profit association having a team of professional data scientists, students, evangelists and enthusiast determined to connect the data science and machine learning community of Australia.
- DSAi: Symbolic Ai Special Edition @ Amazon Web Services
DSAi: Symbolic Ai Special Edition @ Amazon Web Services PLEASE NOTE: that RSVPing to this page DOES NOT GRANT YOU ACCESS to this meetup, Spaces are limited! Due to popular demand & Amazon Facilities Management, PLEASE SIGNUP for A TICKET via the EVENTBRITE LINK BELOW: >>CLICK EVENTBRITE LINK BELOW<< - https://www.eventbrite.com.au/e/dsai-symbolic-ai-special-edition-tickets-58337658507 __________________________________________________ DESCRIPTION Amazon Web Services and the Data Science and Ai Association of Australia presents our Symbolic Ai special edition event. Come along to learn from renowned industry leaders on the practical limits on what computers can and cannot do & how a million-dollar challenge might be solved using Symbolic Ai and Computational complexity theory. ~ Professor Claude Sammut Combining Symbolic and Sub-Symbolic AI for Robot Learning Machine learning has had many recent successes, but current popular methods have their limitations, including requiring many training examples and a lack of transparency. Classical AI techniques can help to learn from a small number of examples and make the result more explainable. In particular, classical AI has well-developed methods for representing and reasoning about background domain knowledge. In this talk, we describe the hybrid learning methods we have used in robot learning that combine symbolic learning and planning with numerical methods. Examples of such robot learning systems include learning a bipedal walk, a robot learning to use tools and a rescue robot learning to traverse rough terrain. Professor Maurice Pagnucco, the Head of the School of Computer Science and Engineering at UNSW will talk about computational complexity theory - one of the most popular topics in computer science today. Computational complexity theory focuses on classifying computational problems according to their inherent difficulty and relating these classes of problems to each other. Did you know that computational complexity theory can help you solve one of 7 Millennium Prize institute Problems? A correct solution to any of the problems will result in a US$1 million prize being awarded by the institute to the discoverer(s). ___________________________________________________________ Biography of presenters Professor Maurice Pagnucco Professor Maurice Pagnucco is Deputy Dean (Education), Engineering and the Head of the School of Computer Science and Engineering at UNSW. Maurice graduated with a BSc (Hons I) and PhD from the University of Sydney. He has held academic positions at the University of Sydney, Macquarie University, the University of Toronto and the University of New South Wales. His research interests lie in: Knowledge Representation and Reasoning Cognitive Robotics Belief Change Reasoning About Actions Claude Sammut is a Professor in the School of Computer Science and Engineering, University of New South Wales. His did pioneering work in Behavioural Cloning. His current interests are focussed on Machine Learning applications in Robotics. He was the leader of the UNSW teams that won RoboCup four-legged robot competitions in 2000, 2001 and 2003; the winners of the 2014 and 2015 Standard Platform League and the team that won the award for best autonomous robot at RoboCup Rescue 2009, 2010 and 2011. In 2012, he was elected to the board of trustees of the RoboCup Federation having previously served as a member of the executive committee from 2003 to 2009. He is co-editor-in-chief of Springer's Encyclopedia of Machine Learning and Data Mining and general chair of RoboCup 2019, Sydney. Schedule 6:00 - 6:30: Beer, Pizza, Networking 6:30 - 6:40: Intro 6:40 - 7:10: Prof. Maurice Pagnucco Talk 7:10 - 7:20: Small break 7:20 - 7:50 Prof. Claude Sammut Talk 7:50 - 8:15 QnA 8:15 - 8:30 Wrap up
- DSAi: ML Ethics & Interpretability Special Edition @ Microsoft
PLEASE NOTE: that RSVPing to this page DOES NOT GRANT YOU ACCESS to this meetup, Spaces are limited! Due to popular demand & Microsoft Facilities Management, PLEASE SIGNUP for A TICKET via the EVENTBRITE LINK BELOW: >>CLICK EVENTBRITE LINK BELOW<< - https://www.eventbrite.com.au/e/dsai-ml-ethics-interpretability-special-edition-tickets-56079112134 ~~~~ DESCRIPTION How do we design Ai systems that we trust? Algorithmic Bias, Algorithmic Transparency, Technological Unemployment, Data Privacy & Algorithmic Misinformation (fake news) are just some of the issues facing the fair and ethical use of Machine Learning. In collaboration with Microsoft for this DSAi special edition Ethics & Interpretability event - come along to learn from industry leaders how issues such as Algorithmic Bias might affect you & what is being done to address the ethical use of Machine Learning in 2019. ~~~~~~~ ‘Ethics for Artificial Intelligence’ In this 20 minute presentation, Aurelie will provide a formal introduction as to what ethical and responsible AI is. Aurelie will go through the current ethical issues that have already appeared using algorithms in society and explain why ethics will play an essential role in successfully designing, deploying and using Artificial Intelligence in the years ahead. Biography: Aurelie is a member of ‘The Australian AI Working Group’, established by Standards Australia with the aim to become the preeminent forum for exploration and discussion of AI by engaging both government and industry; using the global IEEE s’ working group for Standard P7000 ‘Model Process for Addressing Ethical Concerns During System Design’; and The European AI Alliance, a forum established by the European Commission. Separately from her initiatives around Ethics and AI, she is a practising lawyer with over 10 years’ experience in Financial Services. ~~~~~ Interpretable Machine Learning Machine learning models have been getting more accurate, but also more complex over the past few years. However in many settings data scientists are often tethered to linear models or decision trees because they are easy (relatively!) to explain. Moreover developments in data ethics and governance are increasing the pressure on data scientists to explain their models, and to ensure that discrimination or other unwanted outcomes are avoided. In this talk, Anthony will outline the latest and greatest in machine learning interpretability and explain why it is a crucial part of any data scientist’s toolkit. Biography: Anthony Tockar is director and cofounder at Verge Labs, an AI company focused on the applied side of machine learning. A jack-of-all-trades, he has worked on problems across insurance, technology, telecommunications, loyalty, sports betting and even neuroscience. He qualified as an actuary, then moved into data science, completing an MS in Analytics at the prestigious Northwestern University. After hitting the headlines with his posts on data privacy at Neustar, he returned to Sydney to practice as a data scientist and cofounded the Minerva Collective, a not-for-profit focused on using data for social good, as well as multiple meetup groups. His key missions are to extend the reach and impact of data science to help people, and to assist Australian businesses to become more data driven. ~ Ethics & Interpretability Fire-Side Chat We believe that Machine Learning Ethics is Interpretability is one of the grand issues to be effectively tackled in the coming decade. To explore ML Ethics & Interpretability further, after the speakers have completed delivering their presentations we will open the floor to a “fire-side” chat interview where the audience is invited to ask questions open to debate. ~ Schedule 6:00 - 6:30: Beer, Pizza, Networking 6:30 - 6:40: Intro 6:40 - 7:10: Aurelie's Talk 7:10 - 7:20: Small break 7:20 - 7:50 Anthony's Talk 7:50 - 8:15 Fire Side Chat 8:15 - 8:30 Wrap up
- Inviting the next generation of SML organisers and volunteers
UPDATE: moved to Wednesday the 30th of Jan. ~~~~ Are you interested in organising events or volunteering for the next generation of Sydney Machine Learning? We have many initiatives lined up for 2019, including workshops & interesting talks from top in their field speakers. We are inviting all those with an interest in helping with events to a roundtable meeting. No worries if you don't have experience or technical expertise; we are just as happy with helpful suggestions and the passion to contribute :) See you there!
- StarAi: Starcraft 2 Machine Learning Environment Special Edition @ Microsoft
PLEASE NOTE: that RSVPing to this page DOES NOT GRANT YOU ACCESS to this meetup, Spaces are limited! Due to popular demand & Microsoft Facilities Management, PLEASE SIGNUP for A TICKET via the EVENTBRITE LINK BELOW: >>CLICK EVENTBRITE LINK BELOW<< - https://www.eventbrite.com.au/e/starai-starcraft-2-pysc2-machine-learning-environment-api-special-edition-tickets-51844337810 PLEASE NOTE: that RSVPing to this page DOES NOT GRANT YOU ACCESS to this meetup, please signup via the eventbrite link ABOVE… ~~~~~~~~~~~~~~~~~ Relative to previous Ai milestones like Chess or Go, complex video games like Starcraft 2 start to capture the messiness and continuous nature of the real world. It is for this reason that large Machine Learning Research efforts are currently underway by Facebook, Deepmind & Tencent to be the first to solve the “Starcraft Problem”. In August 2017, Blizzard & Deepmind released the PySc2 API in order to perform cutting edge Machine Learning research in the Starcraft 2 environment. On the 6th of November, SML & StarAi is lucky enough to have Steven Brown in Sydney for one night only. Steven is the second greatest contributor to the PySc2 API & author of just about every useful blog post out there on how to build an intelligent Starcraft 2 agent. In partnership with the Microsoft Reactor, StarAi is opening its doors for the final StarAi lecture session this year. Come along to learn about the PySc2 API & how to build a basic Q learning bot in Starcraft 2. Pizza will also be provided. We will also be presenting prizes to participating students who contributed the most throughout the 7 weeks. Requirements: -please bring your laptop, fully charged if possible We only have 60 spots for this event, so get yours now. ================= You may be photographed or videod while attending this event: please approach the organiser if you have questions, concerns, or do not wish this to happen. The event organisers may add you to a mailing list so they can follow up, and you can receive updates about their activities: this mailing list will have an easy-to-find unsubscribe button.
- Sydney Machine Learning Startups Spark Conference
PLEASE NOTE: that RSVPing to this page DOES NOT GRANT YOU ACCESS to this meetup, Spaces are limited! Due to popular demand & Amazon Facilities Management, PLEASE SIGNUP for A TICKET via the EVENTBRITE LINK BELOW: >>CLICK EVENTBRITE LINK BELOW<< - https://www.eventbrite.com.au/e/sydney-machine-learning-startups-spark-conference-tickets-49160943706 PLEASE NOTE: that RSVPing to this page DOES NOT GRANT YOU ACCESS to this meetup, please signup via the eventbrite link ABOVE... ___________________________________ DESCRIPTION In partnership with Spark Festival & Amazon, Sydney Machine Learning is proud to announce the Sydney Machine Learning Startups Conference. Come by at any time between 1pm and 7pm for beer, pizza, networking and speakers, presenting throughout the day on all things Artificial Intelligence / Machine Learning (AI/ML) with a particular focus on the startup ecosystem. You will hear from experts from AI/ML startups discussing their journey and use cases along with some key industry trends. In addition, to drive awareness of deep learning and the role of AI in autonomous driving, Amazon will host the Robocar Rally, where several teams will compete in training a 1/16 scale RC car to drive itself autonomously around a 70sqm track outside the conference rooms. Schedule: 1pm - Jacky Koh, Vylar - Many businesses invest millions in building a data science team yet is still failing to deliver any noticeable value using data science and AI. Jacky will explore why these businesses are failing at ai and how tools like automated machine learning can help companies bootstrap ai. 2pm - Malcolm Layton, BitWoke - has developed a secure Edge AI platform for billions of IoT devices so product innovators can provide real-time AI anywhere. 2:30 pm - Christine Lion, FluroSat - Satellite data and machine learning are revolutionising agriculture. Modern agronomists and farmers need smart algorithms to become more proactive in crop management. FluroSat is exploring several ML applications in its decision support tool FluroSense - from analysis of satellite imagery to training models on agronomic observations. 3pm - Paul Conyngham & Ross McIntyre, WattBlock - Building services segregation using Deep Learning. Learn how Wattblock is applying Deep Learning & other machine learning techniques to generate real & actionable insights into building services seggregation using metering data. 4pm - Beth Carey, Pat Inc - Building a machine intelligence startup. Starting in Sydney, with R&D around Natural Language Understanding, to a Deep technology US company in Silicon Valley Beth talks about her company's journey. 4:30pm - Alex Moss, Canaria Technologies - Predictive Bio-metric Systems: Why to Build? How to Build? The talk will cover multiple problems that can be solved with a decent ECG dataset (ie Epilepsy prediction, consumer fitness tracking) and some of the best techniques to use for each of these. 5pm - Sam Zheng, CuriousThing.io - Conversational AI, opportunities and challenges. 6pm - Tim Garnsey, Verge Labs - Timing the market on AI. Every day we hear stories of success in pushing the theoretical boundaries of AI, yet in the real world the applications stay stubbornly stuck in small niches. This talk covers the current state of what works well in AI today, where to expect new advances, and what parts will be confined to the lab for a very long time. 6:30pm - Andy Huang, Servian - Data, Cloud & AI, Servian We reserve the right to modify the schedule & list of speakers without notice. Spaces are limited, so get your tickets now!
- StarAi Deep Reinforcement Learning Course
StarAi enrolement is now closed. ______ For more information on this initiative, please follow the link below: https://sites.google.com/view/starai-course-beta/home ______ Deep Reinforcement Learning has been responsible for some of the most recent breakthroughs in Machine Learning, from AlphaGo to Dota but also has tremendous commercial application potential. From reducing Google’s data centres cooling costs by 40% to autonomous drone navigation, the applications of RL will continue to increase. We are extremely proud to announce the release of our StarAi deep reinforcement learning course to the general public. Commencing in early September, the course will run for approximately 7 weeks, once a week after work hours from 6 - 8pm. Attendees of the study group can expect to learn: The Epsilon Greedy Algorithm Markov Decision Processes & Dynamic Programming Monto Carlo Sampling, Temporal Difference Learning Tabular Q methods Deepminds DQN Algorithm Policy Gradient Algorithms Proximial Policy Optimization Deepmind’s Pysc2 Starcraft 2 API. What we’re looking for You are comfortable with basic linear algebra. It’s fine if you have to brush up on these skills before course commencement. You are comfortable programming in Python (other languages are helpful, but you’ll spend the program writing in Python). You are comfortable with Git & basic command line calls. eg git push. We’ll use these criteria for selection: Impact on you. We want to understand why this course will help you achieve something you couldn’t otherwise. Self-motivation & communication. We’re looking for people who will work hard through those 7 weeks, and who will inspire others (in the course and externally) to endeavor to learn reinforcement learning as well. Technical skills. The stronger your technical background, the more time you’ll spend focusing on the reinforcement learning itself. More information will be emailed out to those who are successful in their application to the course, Best of luck & regards, The StarAi team.
- Sydney Machine Learning, Tues, 4th of Sep @ Microsoft
Microsoft will be our host for this Sydney Machine Learning meetup at Cliftons (Level 3, 10 Spring St) in Sydney CBD. Come along for beer, pizza, networking and talks from Matt Gibson, Ph.D. student working in computer vision at UNSW’s machine learning group, Kosuke Fujimoto from Microsoft’s Tokyo based Machine Learning group & Krit Kamtuo Software Engineer at Microsoft APAC for Machine Vision focused talks. First Presentation Summary: Title: Learning with not quite the right labels Summary: Having the right data is often as important as having the right model for building effective computer vision systems. Even if you have enough data, what do you do when don't have the right labels? Common problems arising in image understanding include only having partially labelled data, have the wrong type of labels and noisy labels. Broadly these problems can be addressed by an area of research called weakly supervised learning. I will give an overview of some strategies coming from weakly supervised learning to help train good predictive models in the presence of these difficulties. The tools used will include ConvNets combined with some interesting applications of familiar friends such as the EM algorithm and graph-based clustering. Matt Gibson is a Ph.D. student working in computer vision and machine learning at UNSW. He is interested in building high performance machine learning systems for visual understanding. Second Presentation Summary: Title: How can we innovate our workstyle at construction sites with Image Classification? Kosuke, Takenaka Corporation is one of the top 5 largest construction companies in Japan. Their site supervisors perform site visits, record the progress of ongoing construction as well as assign appropriate resourcing from their workforce by writing up the report every evening. This involves taking a lot of photos and organizing them based on the construction progress. In this talk, I walk you through an effort by Takenaka with help from Microsoft to automatically organize photos utilizing Keras and Transfer Learning to make workstyle innovation at construction sites possible. Third Presentation Summary: Title : Using Batch AI to train NLP based on TensorFlow and CNTK in the retail industry Summary: Currently, SSG.COM, the subsidiary of Shinsegae Group the largest retailer in South Korea, is developing an AI chatbot service based on deep learning. It was decided that the internet as as service-based architecture is not efficient because it requires additional resources if the operation production scale is increased. We will discuss why the AI model training layer does not have to be running at all times but the GPU-supported VM should be used continuously, to reduce costs to SSG.COM. Openhack This meetup is a collaboration with Microsoft’s Openhack which will be happening from the 4th to the 6th of September. For more info, you can find more about it here: https://www.microsoftevents.com/profile/form/index.cfm?PKformID=0x43738390001 Time: 6:00 – 8:30 pm Venue: cliftons, level 3, 10 spring street sydney Schedule: 18:00 -18:30 – dinner, drinks & network 18:30 – 19:05 – Matt Gibson’s talk 19:05– 19:15 – break 19:15 – 20:00 – Kosuke & Krit’s talks.
- SML: Deep Reinforcement Learning Special Edition @ Amazon
PLEASE NOTE: that RSVPing to this page DOES NOT GRANT YOU ACCESS to this meetup, Spaces are limited! Due to popular demand & Amazon Facilities Management, PLEASE SIGNUP for A TICKET via the EVENTBRITE LINK BELOW: >>CLICK EVENTBRITE LINK BELOW<< - https://www.eventbrite.com.au/e/sml-deep-reinforcement-learning-special-edition-amazon-tickets-47978801888 PLEASE NOTE: that RSVPing to this page DOES NOT GRANT YOU ACCESS to this meetup, please signup via the eventbrite link ABOVE... ___________________________________ DESCRIPTION Sydney Machine Learning is lucky enough to have Amazon as our host for this special edition Deep Reinforcement learning meetup. Come along for beer, pizza, networking and talks from Alex Long PhD Candidate in the AI group at UNSW focusing on Deep Reinforcement Learning & Alasdair Hamilton CEO of Remi.ai for two extremely interesting talks on Reinforcement Learning. Deep Reinforcement Learning is behind the recent success of ML algorithms being able to learn how to play Atari 2600 games from raw pixel outputs & also AlphaGo, Deepminds machine learning system that overthrew Go world champions, including Lee Sedol in 2016. This session of talks will focus both on the practical side of RL, with Alasdair’s talk focusing on where RL can be actively applied within your business & Alex’s talk taking us into a technical dive of modern Deep Reinforcement Learning techniques. Sydney Machine Learning will also be announcing something really special with regards to Reinforcement Learning which we have been working on since August last year, so if you would like to reserve your spot and find out more, see you on the 8th ;) Alasdair Hamiltons Talk: The future of Artificial Intelligence is, at least partly, rooted in Reinforcement Learning. During our presentation, the team from Remi AI will take attendees on a journey through their experiences in Reinforcement Learning. Special attention will be paid to the areas in which the team have been able to apply Reinforcement Learning in a commercial setting and the four mammalian methods of learning, and how they inspire us. Remi AI has applied RL to budget and bid management, to dynamic pricing and web design, to large scale management of power requirements, inventory management, predictive maintenance. The talk will conclude with a discussion on future applications, as well as a roadmap for those looking to start out in the space, then Q&A. Alasdair Hamilton has cemented himself as a voice of authority in A.I in Australia. Since first encountering the concepts while studying philosophy in 2013, Alasdair has built a successful A.I research firm (Remi AI) that constantly pushes the boundaries of A.I methodologies whilst also applying said methods in real businesses solving real problems. Wearing both CEO and CTO hats, Alasdair knows what is required to run a successful A.I business, bringing multiple platforms to market, in a staggeringly wide number of fields. Alex Longs Talk: Deep Reinforcement Learning (Deep-RL) is the combination of traditional RL algorithms with the high-dimensional function approximation methods of deep learning. This combination allows Deep-RL to eclipse human performance on systems of previously intractable state-spaces and high branching factors, such as the game of GO, Atari arcade games, and heads up limit poker. In this talk I will focus on the intuition behind Deep-RL, how it compares (and differs) to other machine learning methods, as well as discuss some potential commercial applications. Alex Long is a Computer Science PhD Student at the University of New South Wales (UNSW) supervised by Alan Blair. His primary research topic is Deep Reinforcement Learning, specifically the application of RL methods to the NLP domain. Prior to UNSW Alex studied an MSc in Electrical Engineering at the Institute for Cognitive Systems, within the Technische Universität München (TUM) where he published work on artificial skin for humanoid robotics.