TechTalks.Tv beta

Invited talks from ICML 2012 conference

This workshop concerns analysis and prediction of complex data such as objects, functions and structures. It aims to discuss various ways to extend machine learning and statistical inference to these data and especially to complex outputs prediction. A special attention will be paid to operator-valued kernels and tools for prediction in infinite dimensional space.

This workshop studies the interactions between algorithms that learn a model, and algorithms that use the resulting model parameters for inference. These interactions are studied from two perspectives.

The first perspective studies how the choice of an inference algorithm influences the parameters the model ultimately learns. For example, many parameter estimation algorithms require inference as a subroutine. Consequently, when we are faced with models for which exact inference is expensive, we must use an approximation instead: MCMC sampling, belief propagation, beam-search, etc. On some problems these approximations yield superior models, yet on others, they fail catastrophically. We invite studies that analyze (both empirically and theoretically) the impact of approximate inference on model learning. How does approximate inference alter the learning objective? Affect generalization? Influence convergence properties? Further, does the behavior of inference change as learning continues to improve the quality of the model?

A second perspective from which we study these interactions is by considering how the learning objective and model parameters can impact both the quality and performance of inference during “test time.” These unconventional approaches to learning combine generalization to unseen data with other desiderata such as fast inference. For example, work in structured cascades learns model for which greedy, efficient inference can be performed at test time while still maintaining accuracy guarantees. Similarly, there has been work that learns operators for efficient search-based inference. There has also been work that incorporates resource constraints on running time and memory into the learning objective.

This workshop brings together practitioners from different fields (information extraction, machine vision, natural language processing, computational biology, etc.) in order to study a unified framework for understanding and formalizing the interactions between learning and inference. The following is a partial list of relevant keywords for the workshop:

• learning with approximate inference
• cost-aware learning
• learning sparse structures
• pseudo-likelihood training
• contrastive divergence
• piecewise training
• coarse to fine learning and inference
• scoring matching
• stochastic approximation
• incremental gradient methods
and more ...

In this workshop we consider the question of how we can learn meaningful and useful representations of the data. There has been a great deal of recent work on this topic, much of it emerging from researchers interested in training deep architectures. Deep learning methods such as deep belief networks, sparse coding-based methods, convolutional networks, and deep Boltzmann machines, have shown promise as a means of learning invariant representations of data and have already been successfully applied to a variety of tasks in computer vision, audio processing, natural language processing, information retrieval, and robotics. Bayesian nonparametric methods and other hierarchical graphical model-based approaches have also been recently shown the ability to learn rich representations of data.
By bringing together researchers with diverse expertise and perspectives but who are all interested in the question of how to learn data representations, we will explore the challenges and promising directions for future research in this area.

The goal of this challenge is to build an algorithm that learns efficiently a policy to serve news articles on a web site. At each iteration of the evaluation process, you will be asked to pick an article from a list given a visitor (136 binary features + a timestamp). To build a smart algorithm, you might want to balance carefully exploration and exploitation and pay close attention to the “age” of the news articles (among other things of course). A quick look on the leaderboard is enough to figure out why that last point matters. It is the overall CTR (click through rate) of your algorithm that will be taken into account to rank it on the leaderboard.

The 25th Conference on Learning Theory (COLT 2012) was held in Edinburgh, Scotland, on June 25–June 27, 2012. COLT is supported by the Association for Computational Learning (ACL).

Plenary talks presented at IEEE ICRA 2012.

None

Consider a robot that is to act as a household assistant in an unknown kitchen environment. This robot has to acquire and use knowledge about where the task-relevant objects, such as the dish- washer and the oven are and how the robot can act on them. A recent advent of smart devices (e.g. smart phones) and high-quality-low-cost sensors (e.g. Kinect) provides us with the aordable resources for the robot which link sensory information to the robot's knowledge base and high-level deliberative components. Resources like this allow the general-purpose service robots to e.g. query information from world wide web, seek help from remote experts through shared autonomy interfaces and to act independently and safely in human living envi- ronments.

In this hands-on workshop we will identify key problems and so- lutions by narrowing down the denition of semantics, we will dis- cuss what is the representative end world model as a result of se- mantic mapping, single out the optimal sensors, consider static vs. dynamic aspects of environment modeling and nally address the life- long learning in order to leverage not only the sensor data but also from human living patterns and behaviors. The workshop will feature excellent talks from researchers from academia as well as industry, live demonstrations, poster session and a working session with an aim to standardize some fundamental concepts in semantic mapping. We plan to build upon the series of related events at previous IROS, ICRA and RSS conferences.

In order to boost the practical use of robot technology not only in industrial settings, more sophisticated robotic solutions have to be elaborated, particularly in terms of autonomy, exibility, interactiv- ity and cooperating with human, ease of use, and safety. In order to be able to develop applications on the short-term and to maintain ecient improvement of European robotics in the long term, a bet- ter cooperation and technological know-how transfer between robot manufacturers and research institutions is essential.

ECHORD (European Clearing House for Open Robotics Devel- opment, FP7-ICT-231143, http://www.echord.info) is an innovative framework aiming at intensifying this collaboration by carrying more than 50 small sub-projects (socalled experiments with specic research foci and scenarios) with consortia composed of academia and indus- try. The whole project is coordinated by the Technische Universitt Mnchen (Germany), University of Naples (Italy), and University of Coimbra (Portugal). This workshop is composed of two parts:

1. A presentation session where an overview of the ECHORD exper- iments will be given by the coordinating partners of ECHORD, then (intermediate) results of the experiments targeted to an in- ternational audience will be presented, followed by discussions.
2. An open discussion session about innovative solutions in and outside ECHORD, future impacts, new applications, limitations and possible improvements, as well as safety concepts.