A week of presentations

Yesterday I have started new course in the Big Data Technologies. The main focus is to give students hands on experience with the most popular Big Data framework Apache Hadoop. Students will solve basic problems from information retrieval on the Wikipedia data.

Today I am giving a presentation in a 5G course about Cloud Computing. The 5G RN is planned for 2020 and will bring tremendous improvement is speed and latency allowing machine-to-machine (M2M) communication, including remote control (I2M). This will be a driving platform for real time Internet of Things (IoT).

Tomorrow I will present our activities in eClub in a conference at the Ministry of Foreign Affairs. The goal of the meeting is to enable the foreign industry university cooperation. Recently we have been very lucky to find local industrial partners requiring solving challenging problems and we have established great cooperation and I will write a separate blog soon.

On Saturday I am joining the New Media Inspiration conference, which  is about the digital lifestyle and is bringing together experts from many different areas. I am joining with a presentation about speech recognition, my old topic.

At first glimpse the topics may seem to be very incoherent, but in my view they pretty well show the latest trends. The Cloud Computing, Big Data and analytics are pushing ahead new kind of applications, such as the personal assistants (Siri, Google Now, Cortana). They all try to learn the human behavior and help people in everyday activities. They live on the most personal and most widespread device the smartphone. The are implemented as a SaaS in the cloud and people communicate with them by voice, the most natural way. All this is running on wireless networks. And the business part? The innovation, future, smart applications are created by independent developers or startups and this is a where the eClub is coming in play. We try to find motivated people and students put them together and let them work on interesting problems. Industrial partners are helping us with keeping focus as well as with funding.

Yes, we are looking for new students and startups. If you are interested to join us in eClub let me know.

Angry Birds in eClub

Radim and Karel students of the CTU won the ECAI Angry Birds AI Competition. They developed the winning program during the eClub Summer Camp. Thanks to all partners Seznam.cz, CTU MediaLab, CIG, Datalab and all participating students and mentors.

This week the best AI researchers met in Prague at the European Conference on Artificial Intelligence ECAI 2014. As part of the conference the organizers prepared an interesting competition Angry Birds. The goal of this game is to shoot birds to destroy pigs. Teams of researchers were competing in designing an intelligent agent, computer program playing the game without human intervention, ideally better than humans. At the beginning developers are provided with a sample program with components returning the scene screenshots, simple trajectory planner and playing control component. The trick is deciding how and where to shoot birds to kill all pigs. The agent first searches the scene for pigs and then estimates possible targets and finds the trajectories to kill them all. Sounds simple? Not at all! The pigs are hiding behind different objects and to choose the best shot, the best strategy is very difficult. Radim and Karel decided to generalize several strategies solving the game hurdles in all levels. The agent estimated which strategy to play to achieve the overall damage. It took quite some time but at the end they have designed the best agent, they are the winners. Great job! Needless to say,  they were very lucky working with a great mentor Tomas in the eClub Summer Camp.

Angry Birds is just a game, but the technology behind is very useful for solving real problems and the lessons learned will be useful.

There are about twenty students working in the eClub Summer Camp on different AI mainly machine learning problems and mobile applications design. Check our web pages to learn more.

ESC Jump Start Meeting

Students met with the mentors at the first eClub Summer Camp Jump Start Meeting yesterday. This was the unofficial ESC opening. The real start is on July 14. Here is a short report for those of you who have missed the first meeting.

Let me first say thank you to all students and mentors who spent almost the whole morning discussing various projects and introducing their ideas.

What was the program? At the beginning I have briefly explained the organization and here is a list of the main milestones:

• Chose the project and its name.
• Sign a CVUT MediaLab Foundation (our sponsor) contract.
• Sign up on the contact board.
• Set the project web page - include eClub and CVUT MediaLab logos.
• Write an essay.
• To access the computational resources sign up to Metacentrum (Vondra).
• If you need an access to data in Metacentrum sign a NDA.
• Prepare the state of the art technology overview.
• Select the best approach, solution.
• Program, run or copy the selected approach.
• Finish and improve the solution.
• Write a final report.

My presentation was followed by an introduction of mentors. They gave short presentations describing the topics of their interest:

• Twitter Translation
• Cloud projects
• Large Scale Hierarchical Classification
• Rest Search Project
• DHC user scenarions
• New Car UI
• Support Vector Machine
• Machine learning at scale
• General Game Playing for Atari 2600
• Scalable predictive models
• Predictor factory
• Angry birds

At the end we had a long discussion and the most frequent topic was what application to chose and who is the mentor. This is the main job for you now: select the project and choose its name. To help you see the list of the ESC Projects Proposals. You can also review the mentors’ presentations. Select carefully, all mentors are ready to help. We are sure we will do a good job together.  We also hope, that we can convert some of the projects to a BSc or MSc theses.

The other important thing is the scholarship. It will be granted by our sponsor CVUT MediaLab Foundation. It is collecting money from the industry sponsors and it uses the funds to support activities such as ESC. Therefore to get the scholarship all the participants of the ESC need to sign a contract with the Foundation. Currently, I am working on the wording. I need to iron out the last details and as soon as it is finished we will share it with you.

I have also updated the eClub web pages, check it out. I will be step by step adding there all the program, materials, such as the Contract, Non-disclosure Agreement, logos, templates of documents, program of presentations and other stuff, stay tuned.

Yes, it is a lot of work in front of us, this is just the  first meeting, but it was very encouraging. I am looking forward to see great results, mobile and Internet apps etc. Follow us on eClub web pages, Facebook or search for the #CTUECLUB or #CTUESC hashtags.

eClub Summer Camp introductory meeting

The eClub Summer Camp (ESC) preparations continue. We are inviting you for an introductory, jump start meeting June 24th to meet mentors and learn about the projects.

We have busily worked the last weeks on the organizational details of the ESC. We have made a big progress. Because of the organizational reasons, we have to postpone the beginning of the ESC for two weeks. To not disappoint you and to give you a jump start let's meet on 24. June.

The main goal of the jump start meeting is the project discussion and selection. We are very much looking forward to hear about your own ideas, prepare a short 10 min presentation, please. We are also very anxious to introduce our projects in the data section. In plain language we will explain what are the suggested projects about and why we should try to solve them. We want you to be as excited about our project as we are. During the jump start meeting we want to justify the projects importance and explain out what problem they solve.

The jump start meeting will start at 9:00. You will meet with your pears and mentors who will be helping you. We will open with a short personal introduction, then we will give a high level description of the ESC focus and will continue with general discussion. I hope, all mentors will be available throughout the day for appointments, for a discussion. We recommend you you finishing the jump start meeting with some idea about your future project and scheduling a meeting with the appropriate mentors to discuss details. If you have your own project also select a mentor and show the short presentation.

The official kick off is the July 14th. You have the opportunity to use the two weeks for a project selection if you do not chose any project during the first presentation. If you pick up one you can start with the link below the project description and begin working on the first milestone, the current state of the art review.

The ESC will take place in two locations. One will be on the top floor of the Faculty of Civil Engineering building CVUT Dejvice. The other place is the study room of the VIC center in Klokner Institute also Dejvice. In both places we will have the Eduroam and wired Ethernet access. The rooms are air-conditioned.

Let me know if you have questions or suggestions, we are all looking forward to work on interesting high-tech problems and waiting for you great ideas.

eClub Summer Camp Projects

Some of you are already passionate to see the list of eClub Summer Camp (ESC) Data projects. We have many more than we planned. Some of you have not sign up. Fix it, here is the form.

The preliminary listing the projects is quite large. Before you try to select your favorite project you may read this blog to get familiar with the categories. Of course we are open to discuss your own ideas too, just talk to us.

The first category is the Information Retrieval (IR), which contains lot of small problems. All of them are focusing on full text search. For example, learning to rank, interleaving and modeling are contributing to improving the SERP. The rest of the tasks in this category are trying to solve problems related to query expansion.

The Semantic Search section is formulating several problems related to future search engines, which will be able to answer complicated questions such as “How old is Tom Cruise?”. To prepare the right answer we need to be able to find the entities in text or in a query.

We also offer topics in the Ad Selection category. The problem is simple to formulate, but not easy to implement. We need to place to SERP or on a customer web page an ad, which would be most attractive for users. We want to maximize the number of users clicks on an ad, to maximize the income.

The classical problem in machine learning is Spam Detection. The fight still continues in new more twisty incarnations. The problems in this category are focusing on these new forms o f spam.

We have also a large range of problems in the Recommendation System category. The best example? Everybody from the field knows the Netflix Competition with the astonishing first price of 1 million USD. Large number of teams tried to recommend the best movies to viewers. Recommendation is very important for all Internet shop and can help dramatically increase the revenue.

Many applications on today Internet are Mashups and this is the next category. They are put together from simple web services with REST API. Imagine for example an accounting system comparing your performance with similar companies. Another segment in this category is the vertical search. These services can be implemented for example with open source packages such as Lucene and Solr.

If you feel the task description is difficult to understand, difficult to imagine what is the particular problem about and why it is listed, do not be frightened, we will help you to get through it. We will soon announce a meeting to answer your questions and help you with selection. This will be also the ideal place to discuss your own ideas. Follow us on twitter #CTUESC or Facebook. If you want to sign up here is the entry form.

eClub Summer Camp

I am proud to announce the first eClub Summer Camp 2014 for students working in Big Data, Mobile or on their Startups. It will be open from July till middle of September. Join us! Here is the a short summary.

I have ran a small survey in March and I found a considerable number of students is ready to work during the summer on interesting projects. Almost 100 students provided feedback, see the results on graphs. The results inspired me and my partners in organizing the first eClub Summer Camp (ESC). The partners are: Seznam, Samsung, CVUT Medialab, FIT, FEL, CIIRC, IAESTE and I hope, I will add some more. We will start beginning of July in the CVUT Dejvice. The ESC is for students, foreign students, doctoral students, anybody who is interested in building something new and innovative. We will also welcome startups teams with their own interesting projects.

The ESC will have three categories Big Data, Mobile and Startup. The Big Data category includes projects from the following areas: Big Data, Cloud computing, machine learning, artificial intelligence etc. In the Mobile category we want to welcome students or teams focusing on mobile devices especially on the new wave of wearables. We will also add to this category advanced Internet applications. We are looking forward to work with teams of students in the Startups category on entrepreneurial projects. We will propose a list of interesting projects for the Big Data category, if you do not have your own. We expect students in the Mobile and Startup categories to join ESC with their own ideas.

What are the requirements? ESC will operate during the July till middle of September period. You have to make a commitment to work minimum of 6 consecutive weeks in the ESC labs in the Dejvice campus on the agreed project. You will be requested to start with a short, not more than one page long essay explaining your chosen project in general words. Before the end of your stay you will have to complete a report summarizing your work. Do not worry we will tell you how to produce these documents, it is easy. You may join a competition for the best app and startup.

eClub will provide experienced mentors in all categories to help in fast and productive start and get over the initial problems. We will organize interesting presentations about smart and mobile apps. We also prepare the entrepreneurship introductory talks with top experts. Upon signing a contract we will provide a scholarship of eight thousands CZK per month. We will organize introductory and very practical course (duration of 2 hours) for our Big Data cluster. It will support the Big Data processing including Hadoop. We will give you access to large data sets in the order of TB for the Big Data category. We will provide a large selection of interesting problems for the Big Data category including machine learning, text retrieval, document classification, recommender systems, meta search etc. We will provide tools and devices for the mobile category. We will introduce startups to VCs.

If you are interested let us know by filling-in this questionnaire, please. The ESC will be organized with many industrial partners and academic supporters. Stay tuned I will get back with more detail soon. If you have any questions, just ask. Make your summer productive.

FIT & FEL in eClub.

eClub is stepping forward to increase  cooperation between university and industry. The FIT and FEL teams will present their work in Cloud Computing, Big Data and Computational Intelligence. Join us on 11th April at 14:00 in the ERA Svět!

The business competition is getting more and more dependent on data across all industries. To stay competitive managers need to do decisions with the best knowledge of the whole business landscape. Banks, hospitals, GSM providers, surveillance cameras, automotive industry and others are generating tons of data.  How to analyze, how to understand, how to take advantage of the data? Cloud, Big Data processing and analytics are the tools helping us to collect, process and present the information from Internet and other sources. It is inevitable for all companies willing to stay competitive to invest and effectively use these tools. The universities play an important role in researching new ways and algorithms. The FIT and FIT teams and our students are working on many interesting problems from the fields of statistical modeling, data mining, and analytics in general. We want to inform you about what we do and show how to benefit from it.

Historically the cooperation between industry and universities is not great in the Czech Rep. Here are few reasons why we need to change this. Many PhDs want to test their research results in real. Many MSc students are asking to work on practical problems to get better motivation, to get introduced to their future employers. We need to modify the curriculum to reflect the future needs. This can be accomplished only by working more closely with businesses.

The business benefits too! The universities introduce students interested in particular problems to real projects suggested by companies. Very frequently students working on their MSc thesis in a company get familiar with the problem and become employees. Many companies do not have resources or knowledge to focus on some of the more complicated problems, We can become an external R&D. The university staff is a good source of experts, good place to start search for new solutions, to consult, to innovate.

To accomplish all of this we need to start a dialog between universities and industry. It is difficult for a company expert to formulate a good MSc or PhD problem. Faculties are of a great help in translating the problem to a language understandable to students and introducing students the task to make things happen. This requires us getting together more frequently finding a common language.

eClub and our FIT & FEL teams will show the latest research results in the fields such as recommender systems, document classification, sentiment analysis, customer behaviour, infrastructure for cloud and big data etc. We will demonstrate the practical implications of our work in a series of carefully crafted presentations. We will pay special attention to emphasize the practical aspects, keeping the things simple, making them understandable. Our hope is that new ideas will spark and inspire you.  We also hope finding new partners during the follow up networking.

Join us and find new innovative solutions for your businesses. RSVP