All posts by Administrator

RL Workshop

         New Industry and Research Trends

         in AI Frontier and NIPS Conferences

Workshop Information:

When: 14:30 p.m. — 16:30 p.m. December 16th (Sunday)

Location: CSC 3-33, U of A

Price: FREE

Abstract: 2018 AI Frontier and NIPS conferences have just finished. Many new industry and research trends were demonstrated. Dr.Yuxi Li will give a review on the conferences and highlight some inspiring thoughts during this workshop.

Google Form Application:
https://docs.google.com/forms/d/e/1FAIpQLSc66Kc-mPr2IozDR6dobfIAG8jva-ccfiiqFfQB2HzLiJxBdA/viewform?usp=pp_url

Speaker:  Dr. Yuxi Li

  • Founder, attain.ai
  • Author, Deep Reinforcement Learning: https://arxiv.org/abs/1810.06339, (150 pages)
  • Former associate professor in China
  • Former senior data scientist in Boston area in USA
  • Talks about AlphaGo and deep reinforcement learning, including one in MIT and one for AI seminar at the University of Alberta
  • PhD/Postdoc, at AMII, RLAI, Computing Science, University of Alberta

AlbertAI Kaggle Camp

Introduction

Kaggle is an online community of data scientists and machine learners, owned by Google, Inc. Kaggle allows users to find and publish data sets, explore and build models in a web-based data science environment, work with other data scientists and machine learning engineers, and enter competitions to solve data science challenges. Kaggle got its start by offering machine learning competitions and now also offers a public data platform, a cloud-based workbench for data science, and short form AI education.

How Kaggle competitions work?

The competition host prepares the data and a description of the problem. Participants experiment with different techniques and compete against each other to produce the best models. Work is shared publicly through Kaggle Kernels to achieve a better benchmark and to inspire new ideas. Submissions can be made through Kaggle Kernels, through manual upload or using the Kaggle API. For most competitions, submissions are scored immediately (based on their predictive accuracy relative to a hidden solution file) and summarized on a live leaderboard. After the deadline passes, the competition host pays the prize money in exchange for “a worldwide, perpetual, irrevocable and royalty-free license to use the winning Entry”, i.e. the algorithm, software and related intellectual property developed, which is “non-exclusive unless otherwise specified”. Alongside its public competitions, Kaggle also offers private competitions limited to Kaggle’s top participants. Kaggle offers a free tool for data science teachers to run academic machine learning competitions, Kaggle In Class. Kaggle also hosts recruiting competitions in which data scientists compete for a chance to interview at leading data science companies like Facebook, Winton Capital, and Walmart.

About AlbertaAI Kaggle Camp

Thanks to the recent advances in AI research and applications, there is a significantly increasing trend of the needs of AI-related developer/engineer/researcher positions. However, it is not only problematic for people from the academia to engage into the industry but also extremely hard for those non-professionals who are interested in investing time and money to learn and utilize the advancement of AI. We believe one of the best ways of gaining experience and knowledge in a new field is doing a lot of practice.

We utilize one of the most popular data science platform, Kaggle, to provide an opportunity that practitioners can gather, discuss and compete. We’d like to provide this great opportunity for everyone who is interested in landing a position in data science, machine learning and artificial intelligence to practice related skills.

Competition 1

Planet: Understanding the Amazon from Space

https://www.kaggle.com/c/planet-understanding-the-amazon-from-space
Every minute, the world loses an area of forest the size of 48 football fields. And deforestation in the Amazon Basin accounts for the largest share, contributing to reduced biodiversity, habitat loss, climate change, and other devastating effects. But better data about the location of deforestation and human encroachment on forests can help governments and local stakeholders respond more quickly and effectively.

Planet, designer and builder of the world’s largest constellation of Earth-imaging satellites, will soon be collecting daily imagery of the entire land surface of the earth at 3-5 meter resolution. While considerable research has been devoted to tracking changes in forests, it typically depends on coarse-resolution imagery from Landsat (30 meter pixels) or MODIS (250 meter pixels). This limits its effectiveness in areas where small-scale deforestation or forest degradation dominate.

Furthermore, these existing methods generally cannot differentiate between human causes of forest loss and natural causes. Higher resolution imagery has already been shown to be exceptionally good at this, but robust methods have not yet been developed for Planet imagery.

In this competition, Planet and its Brazilian partner SCCON are challenging Kagglers to label satellite image chips with atmospheric conditions and various classes of land cover/land use. Resulting algorithms will help the global community better understand where, how, and why deforestation happens all over the world – and ultimately how to respond.

Competition 2

Quora Question Pairs

https://www.kaggle.com/c/quora-question-pairs

Where else but Quora can a physicist help a chef with a math problem and get cooking tips in return? Quora is a place to gain and share knowledge—about anything. It’s a platform to ask questions and connect with people who contribute unique insights and quality answers. This empowers people to learn from each other and to better understand the world.

Over 100 million people visit Quora every month, so it’s no surprise that many people ask similarly worded questions. Multiple questions with the same intent can cause seekers to spend more time finding the best answer to their question, and make writers feel they need to answer multiple versions of the same question. Quora values canonical questions because they provide a better experience to active seekers and writers, and offer more value to both of these groups in the long term.

Currently, Quora uses a Random Forest model to identify duplicate questions. In this competition, Kagglers are challenged to tackle this natural language processing problem by applying advanced techniques to classify whether question pairs are duplicates or not. Doing so will make it easier to
find high quality answers to questions resulting in an improved experience for Quora writers, seekers, and readers.

Competition 3

https://www.kaggle.com/c/talkingdata-adtracking-fraud-detection

Fraud risk is everywhere, but for companies that advertise online, click fraud can happen at an overwhelming volume, resulting in misleading click data and wasted money. Ad channels can drive up costs by simply clicking on the ad at a large scale. With over 1 billion smart mobile devices in active use every month, China is the largest mobile market in the world and therefore suffers from huge volumes of fraudulent traffic.

TalkingData, China’s largest independent big data service platform, covers over 70% of active mobile devices nationwide. They handle 3 billion clicks per day, of which 90% are potentially fraudulent. Their current approach to prevent click fraud for app developers is to measure the journey of a user’s click across their portfolio, and flag IP addresses who produce lots of clicks, but never end up installing apps. With this information, they’ve built an IP blacklist and device blacklist.

While successful, they want to always be one step ahead of fraudsters and have turned to the Kaggle community for help in further developing their solution. In their 2nd competition with Kaggle, you’re challenged to build an algorithm that predicts whether a user will download an app after clicking a mobile app ad. To support your modeling, they have provided a generous dataset covering approximately 200 million clicks over 4 days!

Stage 1

We will host a two-stage camp. In the first stage, we’d like to invite people with various background to participate in one of the two selected (finished) competition. Kaggle provides late submission options that can allow us to submit predictions and receive the evaluation results immediately. Some guidance will be provided through the one month period. We’d like to select a group of participants to join the next stage. During the first stage, mentorship will be provided by several active Kaggle competitors (Kaggle Master and Kaggle Expert level competitors).

Rules in Stage 1

There is no specific rules for the stage 1, you are very welcome to use any public resources from the Internet. There are sufficient amount of public kernels and competition techinique summaries in the competition discussion forum. Feel free to grab any idea, code and report from others. We don’t pay all the attention to your final performance but your team collaboration and contribution.

Stage 2

In stage 2, selected participants will be invited to form several groups (or not, if someone wants to do a solo competition). We will select a competition from one of the following forms.

  • An ongoing competition
  • A private competition

In stage 2, mentorship will be fully provided to help you reach a higher performance.

Rules in Stage 2

Please strictly follow the rules for the stage 2 listed by Kaggle.

Timeline

Stage 1

  • Begin: Dec 15, 2018
  • End: Jan 15, 2019 (tentative)

Stage 2

  • Begin: Jan 20, 2018 (tentative)
  • End: TBD

How to participate?

We welcome everyone who is interested in spending time on learning and practicing with data science, machine learning and artificial intelligence. However, to ensure the quality of the participants in terms of the engagement and the commitment time, we will charge a $10 entrance fee for every participants for entering the first stage of the camp. This amount will be fully refunded after you finish the first stage.

In addition, a minimal programming skill required for participating is that you should be familiar with basic programming fundamentals (first-year CS course in the university).

Please fill the following Google Form to register to the camp.
https://goo.gl/forms/hrFL466CNl7NfL4k1

In the stage 1, we will gather all the participants and help to form groups based on personal background and interests. Only the first gathering is mandatory for participants to attend. After that, each group can schedule their own preferred meeting time.

Deliverable

Each group is expected to submit the code base and a short report (usually 1 – 2 pages are enough) that demonstrates the efforts of the team. Here we provide a list of previous summary written by top Kagglers.

https://www.kaggle.com/c/quora-question-pairs/discussion/34325

Prize

We will select the top team in the stage 2.

The choice of gift card ($200) will be given as the final prize.

Granular Computing: Pursuing New Frontiers of AI & Data Science

Presented By: Prof. Witold Pedrycz

Abstract:

Artificial Intelligence (AI) has been enjoying a great deal of interest in the recent years owing to its successful applications reported in various areas including robotics, computer vision, and a plethora of systems engaging natural language processing. Data Science becomes instrumental in data analysis discovering essential dependencies and building prediction and classification models.

We advocate that bringing the concepts, methodologies, and formal frameworks of Granular Computing into the area of AI and Data Science can open new and promising avenues of further research and applications. Granular Computing is defined as a coherent and unified platform of acquiring, processing and communication of information granules delivering a highly needed abstraction mechanism.

The talk is self-contained: the principles and formalisms of information granules along with their underlying design principles are discussed in detail. It is shown that the notion of information granularity assumes an important position by bridging the gap between symbolic and sub-symbolic way of processing realized in AI.

The roles of information granules in various faculties of AI such as knowledge representation, concept formation and their hierarchical organization, problem solving mechanisms are elaborated in detail.   It is also emphasized that information granules play a visible role in revealing and characterizing interpretable relationships (dependencies) in data.

When: Tuesday, November 20th, 5:30pm (doors open)

Location: ETLC 1-003 room at University of Alberta, 116 St NW Edmonton, AB T6G 2V4

Registration: https://events.vtools.ieee.org/m/179706

Panel Discussion: Considering having a business in the Artificial Intelligence industry?

– By TusStar Alberta, TEC Edmonton and AlbertaAI

4:00-6:00 PM, Oct. 17, 2018 Wednesday (Mountain/Alberta Time)

Organizers: TusStar Alberta, TEC Edmonton and AlbertaAI

 

Abstract:

Artificial Intelligence (AI) has been an essential part of our daily life. AI has developed a large number of tools to solve the most difficult problems in various areas. For example, the SeeingAI app developed by Microsoft can help people who are blind perceive the environment around them. For the past few years, AI has entered the consciousness of almost every industry. It has become one of the hottest topics of mainstream conversations. TusStar Alberta, TEC Edmonton and AlbertaAI have invited speakers with various backgrounds to give you insights about having a business in the AI industry from different perspectives.

Agenda:

4:00 – 4:30 Registration and networking
4:30 – 5:30 Panel
5:30 – 6:00 Networking

Speakers:

Bruce Alton, Partner of A-Partners

Bruce Alton is the founder of A-Partners which is an investor in, advisor to and founder of high-growth enterprises with a particular focus on companies that utilize artificial intelligence, machine learning and data analytics as a competitive business advantage. 

He has been a company founder and senior operating executive and his thirty-year business career has spanned software development, advanced manufacturing, venture capital, energy and real estate.

Bruce was previously named as one of Alberta’s 50 Most Influential People by Alberta Venture Magazine. He has been recognized for his work with several Alberta community organizations including the A100 – Alberta Technology Entrepreneurs Organization; the Kids With Cancer Society; and the Alberta Economic Development Authority.

 

Dave Damer, CEO of Testfire Labs

Dave Damer is a computer engineer and proven tech entrepreneur with over 27 years of business leadership experience, 18 of which have been spent building startups. 

In 2017, Dave founded Testfire Labs, a software company that uses machine learning and artificial intelligence to build productivity solutions that modernize the way people work, so they can achieve more, faster.

Testfire’s flagship product is Hendrix.ai, an AI-assistant that transcribes meeting summaries and action items, audits meeting history, and drops it all into a simple, searchable dashboard, for unprecedented organization and productivity insights.

Testfire Labs has been named a Top Startup to Watch by Startup Edmonton, received the 2018 Startup Canada Prairies Region Innovation Award, and has been nominated for the Startup of the Year award by the National Angel Capital Organization (NACO).

Dr. Jin Shi, Director of Solutions Architecture, CenturyLink

Dr. Shi Jin has been leading the SafeHaven engineering team for the past 5 years, growing it from a small Edmonton based startup (DataGardens) to be part of one of the largest global technology companies (CenturyLink). 

Before that, Dr. Jin Shi worked in companies like Acrodex and Cybera. He also co-founded Canada’s first infrastructure as a cloud service company VRStorm.  Dr. Jin Shi got his BS from Tsinghua University of China in 2001 and PhD from Cornell University of USA in 2006.

 

Dr. Yuxi Li, PhD & Post-doc of Computing Science department at University of Alberta

Dr. Yuxi Li has more than ten years of experience in reinforcement learning, machine learning, and AI.  

Dr. Yuxi was an associate professor in China, and a senior data scientist in a fintech company in Boston area. He is a co-organizer for AI Frontiers Conference (aifrontiers.com) in Silicon Valley in 2017 and 2018. He serves as a Program Committee Member for AAAI 2019, a top AI conference. He published the influential Deep Reinforcement Learning: An Overview on arXiv, and is preparing for a book, to be available in early 2019. He is the founder of attain.ai.

Wei Sun, Vice President of ChiHe Investment Inc.

ChiHe focuses on policy-supported industries, makes equity investments in private enterprises with potentials, and enhances the financial and industrial resources of small and medium-sized enterprises, to boost the real economy. 

ChiHe make equity investment through private equity funds. By leveraging rich investing experiences, we manage the whole process including investment, financing, integration and exit. While achieving expected returns for fund investors, we provide private enterprises with funds, policy supports, market channel and other resources, to help them expand business in right geographic areas and grow rapidly. In 2016, ChiHe and its partners completed acquisition of Analogix Semiconductor in US.

Click here or scan the following bar-code to register today! 

Talk: Open Data, Data Science and Digital Innovation at City of Edmonton

When: 6:00-9:30 PM, Sep. 12, 2018 Wednesday

Speaker: Norman Mendoza, Ben Gready, Roman Eisne,
Jared Rewerts, Jeanna Somoza,
Smart City & Analytics Centre of Excellence,
City of Edmonton

City of Edmonton is a leading city on Open Data and Smart City. There are dedicated teams in the City focusing on open data, digital innovation, data science and disruptive technology.

Edmonton’s Open Data Strategy integrates the principles of the International Open Data Charter, outlines an action plan to realize the value of open data and provides metrics to evaluate performance. Through the application of this Open Data Strategy, Edmonton will continue to demonstrate international leadership in the practice of open government and will fulfill its role as a preeminent global city.

Data released by the City has the potential to be both impactful and beneficial, contributing to the livability of the city. Open data allows for direct visibility into how decisions are made, empowers citizens and enhances the relationship between citizens and public organizations.

Edmonton’s Open Data Portal provides public access to hundreds of datasets within their Open Data Catalog. Available in multiple formats, they can be downloaded for free.

City of Edmonton is also one of Canada’s Top Employers For Young People. In this event, leaders and a former summer student from the City of Edmonton introduced open data and the work related to data science, smart city and AI. There was an interactive panel discussion after the speeches.


Open Data at the City of Edmonton

(by Norman Mendoza, City of Edmonton)

Overview and Examples of Open Source
and IoT Projects at the City of Edmonton

(by Jared Rewerts, Software Engineer,
Jeanna Somoza, former Summer Student)

Data Science Projects at the City of Edmonton

(by Ben Gready & Roman Eisner, Data Scientists)

Panel Discussion

(by Norman, Ben, Roman, Jared, Jeanna, Lihang Ying, Frank Wang)


Organizers, Partners and Sponsors

Relevance of Human Intuition in AI Driven World

– By Shri Parnerkar Institute for Research, Innovation and Technology Foundation

Presenter: Shri Parnerkar Institute for Research, Innovation and Technology Foundation
Organizer: DPLMF
Co-Organizer: Alberta AI
Outreach Partner: TusStar
Where: Maple Leaf Room, Lister Center, University of Alberta
Google Map: https://goo.gl/maps/1C2sAPTZLWL2

When: 6:00 PM – 9:00 PM on October 24, 2018

A unique think tank discussion with eminent panelists.


The panel consists of industry professionals from various fields such as medicine, artificial intelligence, psychology, entrepreneurship, and more. This interactive seminar is designed to discuss the relevance of human intuition in an artificial intelligence driven world. How can Artificial Intelligence help humanity? How can a common man leverage Human Intuition? What is the role of intuition in Artificial Intelligence/Healthcare/Entrepreneurship? Join us at this one of a kind event and leave with your curiosities satisfied.

Congratulations to Mr Wenfei Zhang for his New Position at Tusstar Alberta

We are thrilled to announce that Dr. Wenfei Zhang, the incumbent VP Strategic Planning of AlbertaAI, has been appointed to the Program Manager of TusStar Alberta on July 23. Congratulations Wenfei!

As AlbertaAI’s strategic partner, TusStar is not unfamiliar to our members. Established in 1999, TusStar is one of the longest-running and most successful Chinese techno-entrepreneurial incubators. With offices in 50+ Chinese cities and 10+ cities overseas, TusStar has provided service to 5,000+ companies, boasting 35 IPOs and over 3 billion dollars in equity return.

TusStar’s Alberta office opened up last June and has attracted much attention and support ever since. The launch of TEC Edmonton-TusStar Accelerator Exchangeprogram on June 11 this year marks its most recent advancement in Alberta as well as in Canada. TEC Edmonton and TusStar are top university incubators in Canada and China, respectively. This six-month program aims to help entrepreneurs in Canada/China to explore the markets in China/Canada.

Dr. Wenfei Zhang is one of the earliest executive team members of AlbertaAI. His service has benefited many AlbertaAI members. Meanwhile, his experience with AlbertaAI also established himself as a recognized contributor to the AI ecosystem in Alberta, which in turn helped him secure the senior position with TusStar Alberta.

It is our belief that the appointment is a win-win for both organizations, and it will set a great example for the career development of more AlbertaAI members. Let’s wish Wenfei greater success in both positions!

Talk: AI and how SAS is delivering an AI platform

Jan 26, 2018
Thanks to all the audiences for our talk: AI and how SAS is delivering an AI platform! There will be more events coming up, follow us on Linkedin or Wechat.
Here follows the slides of this talk,credit to SAS executive—-Steve Holder:
 

AlbertaAI_SAS talk by Steve Holder.pdf

Download File

The TusStar Start-up Incubator Presentation

By Yongzhe Hong
​Nov 28, 2017

We would like to thank the generous sponsorship from Barry Bian (TD Wealth Financial Planner).

 

 

Ms. Wei Han, Director of Institute Collaboration of Tusholdings Co. and Deputy General Manager of TusStar Business Incubator visited Edmonton with a delegation of Tusholdings and interacted with Chinese Immigrants and students on Nov 22, 2017. This event was organized by AlbertaAI Association, University of Alberta Chinese Graduate Students’ Club (UA – CGSC), Edmonton IT Club (EITC) and Tsinghua Alumni Association of Edmonton (TAAE). There were more than 70 people attended this event.

 

 

At the beginning, Frank Wang, the VP-External of AlbertaAI and president of TAAE introduced the main event organizer—AlbertaAI, which is a new non-profit and non-government organization founded by a group of passionate volunteers. This organization is dedicated to contribute to Alberta’s AI eco-system and connect Alberta AI community to the world.

Then, Ms. Han introduced the global innovation network of TusStar business incubator, early investment, eco-system for innovation and entrepreneurship, starting a business in China for oversea talents. Tusholdings is a large comprehensive enterprise strongly supported by Tsinghua University. TusStar, as the flagship company known for its innovative incubation and investment business, was named “Top 10 incubator in China” in 2014. The four-in-one platform of TusStar provides incubation service, open platform, training program and early investment. TusStar not only has established incubator bases in more than 60 cities in China, but also has global SciTech innovation net in 12 countries over 6 continents, greatly reduce the difficulty of overseas development for start-ups. It is worth mentioning that in this June, TusStar has officially opened a branch company in Edmonton, Alberta.

 

Working together with the Edmonton local business incubator “TEC Edmonton”, the two will establish an innovation fund to invest excellent companies. Ms. Han also introduced the AI innovation Centre in Shanghai, China. This innovation Centre will provide corresponding support for start-ups, including financial, legal, patent, tax and other aspects of professional services.

 

 

​In next round, three founders of business projects in the field of AI briefly introduced their project. Dr. Lihang Ying, the president of AlbertaAI introduced a location based instant smart notification app— Boardee, in which he is a cofounder. It is dedicated to the mission of aggregating location-oriented information, understanding personal contextual interests and pushing the right information on the right location and time.

 

 

 

 

Following that, Dr. Ming Dong shared his project of an infrastructure management platform.  This cloud computing products can provide intuitive, convenient, visual equipment health and risk assessment, and provide professional advice for the equipment maintenance, replacement and the selection of specific type.

 

 

 

Finally, Liming Zhang introduced a cross-border mobile payment platform project. This platform is now the authorized partners of Wechat pay and Alipay. In addition, it can also provide information display and shopping platform for Chinese tourists travelling in Canada.

 

 

 

Close to the end of the event, Tusholdings delegation interacted with the audiences and answered questions from them. Today, several audiences drove more than 300 kilometers from Calgary just for this event. The chief operation officer of TusStar Edmonton Branch Company Lixin Liu and TEC Edmonton director Lan Tan also attended today’s event.