Plenary lectures

Accepted papers

Following papers have been accepted for Journal Track of our conference:

Latebreaking track talks

Following talks have been accepted for Latebreaking track:

About

Conference on Theoretical Foundations of Machine Learning (TFML 2015) will take place in Będlewo, Poland, on February 16-21, 2015. We invite submissions of papers addressing theoretical aspects of machine learning and related topics. We strongly support a broad definition of learning theory, including, but not limited to:

• Theoretical foundations of
  • supervised learning
  • unsupervised learning
  • semi-supervised learning
  • deep learning
  • active learning
  • concept drift
  • natural language processing
  • neuroinformatics
• generalization bounds
• optimization methods
• computational complexity of learning
• information theoretic learning

Conference is organized by Department of Machine Learning, Institute of Computer Science and Computational Mathematics, Faculty of Mathematics and Computer Science, Jagiellonian University

Invited Speakers

• Bernhard Geiger
  "Markov State Space Aggregation via the Information Bottleneck Method"
• Eyke Hüllermeier
  "Online preference learning with bandit algorithms"
• Małgorzata Bogdan
  "SLOPE Adaptive Variable Selection via Convex Optimization"
• Manuel Graña
  "Lattice computing for Intelligent Systems"
• Jerzy Ombach
  "A Short Introduction to Stochastic Optimization"

Publication model

• The journal track. Original, unpublished research papers can be submitted to this track through easychair system. They will be peer reviewed and after acceptance and presentation during the TFML conference published in Schedae Informaticae Journal.
• The latebreaking talks track, which is aimed at researchers willing to present their latest work during TFML, but without publishing it in journal/proceedings. This works will still undergo peer review and have to be uploaded to arXiv and registered in the easychair system.

Conference materials

FAQ

1. What is the difference between Journal track and latebreaking track?
   The only difference between those two tracks is a publication model, in each case you submit and present your work at the conference. However, in the journal track your work gets published in Schedae Informaticae Journal, while in latebreaking track you only upload your work to arXiv repository. Both tracks are peer reviewed.
2. Why would I choose latebreaking track instead of Journal one?
   There are two main reasons, first, latebreaking track undergo lighter review process than the journal track, as we also welcome "in progress studies" and similar ones in latebreaking track. Furthermore, ones work might be already under review in some other journal (or even is already accepted) and so it is not possible to publish it twice. Such results would still qualify as a good talk at TFML, while not a good journal track submission.
3. How to submit my work to latebreaking track?
   If you have not done it yet, upload your paper to arXiv repository. Once it is done, click on the "register paper" link in the right column and complete the paper submission form with the .pdf file from your arXiv submission. Remember to add [arXiv] tag to your manuscript title.

Key dates

• paper submission
  Journal track: 10 Nov 10 Dec 2014
  Latebreaking talks track: 10 Dec 2014
• decision notification - 5 8 Jan 2015
• camera ready version - 12 15 Jan 2015
• conference - 16-21 Feb 2015
• registration deadline - 20 Jan 2015
• payment deadline - 20 Jan 2015

Journal track

Submission format

Papers should be prepared in English, according to the Schedae Informaticae template. Page limit (including references, and other additional materials) is 8 pages (soft limit) and a hard limit of 10 pages. Papers longer than 10 pages will be rejected without reviewing.

Submit your paper

EasyChair submission system is now closed. check your paper status

Publication

All papers will be peer-reviewed. All accepted papers will be published in Schedae Informaticae Journal¹. At least one author of each publication has to attend the conference and present it during oral/poster session.

¹For Polish scientists, this journal is located on the B list of the Polish Ministry of Science and Higher Education, with 10pts.
www.ejournals.eu/Schedae-Informaticae/

Latebreaking talks track

If you are not interested in publication you can still submit your talk for TFML and present your work during our conference.

Submission format

Papers should be uploaded to arXiv.

Submit your paper

submit your paper to latebreaking track and register your talk through EasyChair - simply upload arxiv's pdf file through our system and add [arXiV] tag to yours submission's title check your paper status

Confused about publication model? See FAQ section

Fees

Registration fee of 1300 PLN (350 euro) covers:

• presentation slot
• publication of one paper
• accomondation and lodging
• food (breakfast, dinners and suppers)
• coffee breaks
• welcome party
• proceedings

There is possibility of discounts for phd students.

Registration

In order to register for TFML conference you have to:

• Fill in the registration form
• Pay the registration fee

Registration fee

The registration fee of 1300zł is composed of two parts, payed independently to two bank accounts not later than 20 Jan 2015.

1000 zł should be transfered to 300 zł should be transfered to As a title of both transfers use the following template.

Schedule

	Mon	Tue	Wed	Thu	Fri
8:00		Breakfast	Breakfast	Breakfast	Breakfast
9:00		Plenary Lecture 2	Plenary Lecture 4	Plenary Lecture 6	Plenary Lecture 8
10:00		Coffee break	Coffee break	Coffee break	Coffee break
11:00		Session 1	Session 3	Session 5	Plenary Lecture 9
12:00					Closing
13:00		Dinner	Dinner	Dinner	Dinner
14:00	Registration	Plenary Lecture 3	Plenary Lecture 5	Plenary Lecture 7	Bus to Poznań
15:00	Dinner	Coffee break	Coffee break
16:00	Opening	Session 2	Session 4
17:00	Plenary Lecture 1			Social event
18:00	Supper	Supper	Supper
19:00				Gala dinner

Each plenary lecture has 60 min + 10 min discussion time slot.

Plenary lectures

1. Małgorzata Bogdan "SLOPE Adaptive Variable Selection via Convex Optimization"
2. Eyke Hüllermeier "Online preference learning with bandit algorithms"
3. Jerzy Ombach "A Short Introduction to Stochastic Optimization"
4. Manuel Graña "Lattice computing for Intelligent Systems"
5. Michał Woźniak "Data stream classifcation using classifer ensemble"
6. Łukasz Dębowski "Hilberg's Conjecture - a Challenge for Machine Learning"
7. Bernhard Geiger "Markov State Space Aggregation via the Information Bottleneck Method"
8. Krzysztof Dembczyński "Decision-theoretic Approach to Multi-label Classifcation"
9. Jacek Tabor "MDLP approach to clustering", Igor Podolak "Hierarchical Classification of Objects"

Each presentation at oral session has a 25 min + 5 min discussion time slot.

Session 1

1. Modelling influence propagation in social networks
2. Multilinear Filtering Based on a Hierarchical Structure of Covariance Matrices
3. Potential Contour Ensembles

Session 2

1. Incoherent Dictionary Learning For Sparse Representation In Network Anomaly Detection
2. Effectiveness of Unsupervised Training in Deep Learning Neural Networks
3. Analysis of compounds activity concept learned by SVM using robust Jaccard based low-dimensional embedding

Session 3

1. Active Function Cross Entropy Clustering
2. Subspace Memory Clustering
3. Mixture of metrics optimization for machine learning problems

Session 4

1. Cross entropy clustering approach to iris segmentation for biometrics purpose
2. Feature Selection and Classification Pairwise Combinations for High­dimensional Tumour Biomedical Datasets
3. Cross­ entropy based image thresholding
4. Demo of gmum.R package

Session 5

1. Looking for the right time to shift strategy in the exploration/­exploitation dilemma
2. On the consistency of Multithreshold Entropy Linear Classifier
3. Fast optimization of Multithreshold Entropy Linear Classifier
4. Maximum Entropy Random Walks - from theory to practise