ALT & DS 2017 Conference Programme

Saturday, 14th October 2017


18:00 - 20:00 Registration & Welcome Reception

Sunday, 15th October 2017


ALT DS
9:00 - 9:30 Registration
9:30 - 10:00 Conference Opening & Awards
10:00 - 11:00 Invited Talk (ALT+DS): Masashi Sugiyama (AIP)
11:00 - 11:30 Coffee Break
11:30 - 12:30 ALT Session 1: Online Learning (Chair: Eji Takimoto) DS Session 1: Online Learning
New bounds on the price of bandit feedback for mistake-bounded online multiclass learning.
Phil Long
Context-Based Abrupt Change Detection and Adaptation for Categorical Data Streams.
Sarah D'Ettorre, Herna Viktor, Eric Paquet
Efficient tracking of a growing number of experts.
Jaouad Mourtada and Odalric-Ambrym Maillard
A New Adaptive Learning Algorithm and Its Application to Online Malware Detection.
Ngoc Anh Huynh, Wee Keong Ng, Kanishka Ariyapala
Scale-Invariant Unconstrained Online Learning.
Wojciech Kotlowski
Real-Time Validation of Retail Gasoline Prices.
Mondelle Simeon, Howard J. Hamilton
12:30 - 14:00 Lunch
14:00 - 15:00 ALT Session 2: Learning and Probability (Chair: András Györge) DS Session 2: Regression
Learning from networked examples.
Yuyi Wang, Zheng-Chu Guo and Jan Ramon
General Meta-Model Framework for Surrogate-Based Numerical Optimization.
Žiga Lukšič, Jovan Tanevski, Sašo Džeroski, Ljupičo Todorovski
Hypothesis testing on infinite random graphs.
Daniil Ryabko
Evaluation of Different Heuristics for Accommodating Asymmetric Loss Functions in Regression.
Andrei Tolstikov, Frederik Janssen, Johannes Fürnkranz
Boundary Crossing Probabilities for General Exponential Families.
Odalric-Ambrym Maillard
Differentially Private Empirical Risk Minimization with Input Perturbation.
Kazuto Fukuchi, Quang Khai Tran, Jun Sakuma
15:00 - 15:10 Coffee Break
15:10 - 16:30 ALT Session 3: Inductive Inference (Chair: Frank Stephan) DS Session 3: Pattern Mining + Bioinformatics
Erasing Pattern Languages Distinguishable by a Finite Number of Strings.
Fahimeh Bayeh, Ziyuan Gao and Sandra Zilles
Mining Strongly Closed Itemsets from Data Streams.
Daniel Trabold, Tamás Horváth
Automatic Learning from Repetitive Texts.
Rupert Hölzl, Sanjay Jain, Philipp Schlicht, Karen Seidel and Frank Stephan
Extracting Mutually Dependent Multisets.
Natsuki Kiyota, Sho Shimamura, Kouichi Hirata
Learning MSO-definable hypotheses on strings.
Martin Grohe, Christof Löding and Martin Ritzert
LOCANDA: Exploiting Causality in the Reconstruction of Gene Regulatory Networks.
Gianvito Pio, Michelangelo Ceci, Francesca Prisciandaro, Donato Malerba
Normal Forms in Semantic Language Identification.
Timo Kötzing, Martin Schirneck and Karen Seidel
Discovery of Salivary Gland Tumors' Biomarkers via Co-Regularized Sparse-Group Lasso.
Sultan Imangaliyev, Johannes H. Matse, Jan G.M. Bolscher, Ruud H. Brakenhoff, David T.W. Wong, Elisabeth Bloemena, Enno C.I. Veerman, Evgeni Levin
16:30 - 16:40 Coffee Break
16:40 - 18:00 ALT Session 4: Learning and Approximation (Sandra Zilles) DS Session 4: Knowledge Discovery
Tight Bounds on ℓ1 Approximation and Learning of Self-Bounding Functions.
Vitaly Feldman, Pravesh K Kothari and Jan Vondrak
Measuring the Inspiration Rate of Topics in Bibliographic Networks.
Livio Bioglio, Valentina Rho, Ruggero G. Pensa
PAC Learning Depth-3 AC0 Circuits of Bounded Top Fanin.
Ning Ding, Yanli Ren and Dawu Gu
Discovering Minority Sub-clusters and Local Difficulty Factors from Imbalanced Data.
Mateusz Lango, Dariusz Brzezinski, Sebastian Firlik, Jerzy Stefanowski
Relative Error Embeddings of the Gaussian Kernel Distance.
Di Chen and Jeff Phillips
Fusion Techniques for Named Entity Recognition and Word Sense Induction and Disambiguation.
Edmundo-Pavel Soriano-Morales, Julien Ah-Pine, Sabine Loudcher
Minimax rates for cost-sensitive learning on manifolds with approximate nearest neighbours.
Henry Reeve and Gavin Brown

Monday, 16th October 2017


ALT DS
9:00 - 10:00 Invited Talk (ALT): Adam Kalai (Microsoft Research)
10:00 - 10:30 Coffee Break
10:30 - 11:30 ALT Session 5: Query Learning (Chair: Sanjay Jain) Special Session: Takeaki Uno
The Power of Random Counterexamples.
Dana Angluin and Tyler Dohrn
(E. M. Gold Award)
An efficient query learning algorithm for zero-suppressed binary decision diagrams.
Hayato Mizumoto, Shota Todoroki, Diptarama, Ryo Yoshinaka and Ayumi Shinohara
Preference-based Teaching of Unions of Geometric Objects.
Ziyuan Gao, David Kirkpatrick, Christoph Ries, Hans Simon and Sandra Zilles
11:30 - 11:40 Coffee Break
11:40 - 12:40 ALT Session 6: Interactive and Transfer Learning (Chair: Phil Long)
Adaptive Submodularity with Varying Query Sets: An Application to Active Multi-label Learning.
Alan Fern, Robby Goetschalckx, Mandana Hamidi-Haines and Prasad Tadepalli
Dealing with Range Anxiety in Mean Estimation via Statistical Queries.
Vitaly Feldman
Lifelong Learning in Costly Feature Spaces.
Maria-Florina Balcan, Avrim Blum and Vaishnavh Nagarajan
12:40 - 18:00 Lunch & Excursion
18:00 - 20:00 Banquet at Conference Venue

Tuesday, 17th October 2017


ALT DS
9:00 - 10:00 Invited Talk (ALT): Alexander Rakhlin (University of Pennsylvania)
10:00 - 10:30 Coffee Break
10:30 - 11:30 ALT Session 7: Bandit Learning (Chair: Odalric-Ambrym Maillard) DS Session 5: Label Classification
A minimax and asymptotically optimal algorithm for stochastic bandits.
Pierre Menard and Aurélien Garivier
On a New Competence Measure Applied to the Dynamic Selection of Classifiers Ensemble.
Marek Kurzynski, Pawel Trajdos
Structured Best Arm Identification with Fixed Confidence.
Mohammad Mahdi Ajallooeian, Ruitong Huang, Csaba Szepesvári and Martin Müller
Multi-label Classification Using Random Label Subset Selections.
Martin Breskvar, Dragi Kocev, Sašo Džeroski
A Modular Analysis of Adaptive (Non-)Convex Optimization: Optimism, Composite Objectives, and Variational Bounds.
Pooria Joulani, Andras Gyorgy and Csaba Szepesvári
Option Predictive Clustering Trees for Hierarchical Multi-label Classification.
Tomaž Stepišnik Perdih, Aljaž Osojnik, Sašo Džeroski, Dragi Kocev
11:30 - 11:40 Coffee Break
11:40 - 12:40 ALT Session 8: Networks and Matrices (Chair: Vitaly Feldman) DS Session 6: Deep Learning
The Complexity of Explaining Neural Networks Through (group) Invariants.
Danielle Ensign, Scott Neville, Arnab Paul and Suresh Venkatasubramanian
Re-training Deep Neural Networks to Facilitate Boolean Concept Extraction.
Camila González, Eneldo Loza Mencía, Johannes Fürnkranz
On Compressive Ensemble Induced Regularisation: How Close is the Finite Ensemble Precision Matrix from the Infinite Ensemble?
Ata Kaban
An In-Depth Experimental Comparison of RNTNs and CNNs for Sentence Modeling.
Zahra Ahmadi, Marcin Skowron, Aleksandrs Stier, Stefan Kramer
Collaborative Clustering: Sample Complexity and Efficient Algorithms.
Jungseul Ok, Se-Young Yun, Alexandre Proutiere and Rami Mochaourab
12:40 - 14:00 Lunch
14:00 - 15:00 ALT Session 9: Teaching and Testing (Chair: Daniil Ryabko) DS Session 7: Feature Selection + Recommendation System
Non-Adaptive Randomized Algorithm for Group Testing.
Nader Bshouty, Nuha Diab, Shada R. Kawar and Robert J. Shahla
Improving Classification Accuracy by Means of the Sliding Window Method in Consistency-Based Feature Selection.
Adrian Pino Angulo, Kilho Shin
Graph Verification with a Betweenness Oracle.
Mano Vikash Janardhanan
Feature Ranking for Multi-target Regression with Tree Ensemble Methods.
Matej Petkovič, Sašo Džeroski, Dragi Kocev
Specifying a positive threshold function via extremal points.
Vadim Lozin, Igor Razgon, Victor Zamaraev, Elena Zamaraeva and Nikolai Zolotykh
Recommending Collaborative Filtering Algorithms Using Subsampling Landmarkers.
Tiago Cunha, Carlos Soares, André C. P. L. F. de Carvalho
15:00 - 15:10 Coffee Break
15:10 - 16:30 ALT Session 10: Bayesian Techniques (Chair: Wojciech Kotlowski) DS Session 8: Community Detection
Universality of Bayesian mixture predictors.
Daniil Ryabko
Recursive Extraction of Modular Structure from Layered Neural Networks Using Variational Bayes Method.
Chihiro Watanabe, Kaoru Hiramatsu, Kunio Kashino
Soft-Bayes: Prod for Mixtures of Experts with Log-Loss.
Laurent Orseau, Tor Lattimore and Shane Legg
Discovering Hidden Knowledge in Carbon Emissions Data: A Multilayer Network Approach.
Kartikeya Bhardwaj, Hingon Miu, Radu Marculescu
A Strongly Quasiconvex PAC-Bayesian Bound.
Niklas Thiemann, Christian Igel, Olivier Wintenberger and Yevgeny Seldin
Topic Extraction on Twitter Considering Author's Role based on Bipartite Networks.
Takako Hashimoto, Tetsuji Kuboyama, Hiroshi Okamoto, Kilho Shin
Parameter identification in Markov chain choice models.
Arushi Gupta and Daniel Hsu
16:30 - 17:00 Coffee Break
17:00 - 18:00 Invited Talk (DS): Koji Tsuda (Tokyo University)
18:00 - 18:10 Closing