The estimation of linear causal models (also known as structural equation models) from data is a well-known problem which has received much attention in the past. Most previous work has, however, made an explicit or implicit assumption of gaussianity, limiting the identifiability of the models. We have recently shown (Shimizu et al, 2005; Hoyer et al, 2006) that for non-gaussian distributions the full causal model can be estimated in the no hidden variables case. In this contribution, we discuss the estimation of the model when confounding latent variables are present. Although in this case uniqueness is no longer guaranteed, there is at most a finite set of models which can fit the data. We develop an algorithm for estimating this set, and describe numerical simulations which confirm the theoretical arguments and demonstrate the practical viability of the approach. Full Matlab code is provided for all simulations.

Abstract-- This article proposes the first known algorithm that achieves a constant-factor approximation of the minim um length tour for a Dubins' vehicle through n points on the plane. By Dubins' vehicle, we mean a vehicle constrained to move at constant speed along paths with bounded curvature without reversing direction. The Traveling Salesperson Problem (TSP) with its variations is one of the most widely known combinatorial optimization problems. While extensively studied in the literature, these problems continue to attract great inter est from a wide range of fields, including Operations Research, Mathematics and Computer Science. It is quite natural to formulate this problem in context of Dubins' vehicle, i.e., a non-holonomic vehicl e that is constrained to move along paths of bounded curvature, without reversing direction.

We discuss here constraint programming (CP) by using a proof-theoretic perspective. To this end we identify three levels of abstraction. Each level sheds light on the essence of CP. In particular, the highest level allows us to bring CP closer to the computation as deduction paradigm. At the middle level we can explain various constraint propagation algorithms. Finally, at the lowest level we can address the issue of automatic generation and optimization of the constraint propagation algorithms.

We study the complexity and expressive power of conjunctive queries over unranked labeled trees represented using a variety of structure relations such as ``child'', ``descendant'', and ``following'' as well as unary relations for node labels. We establish a framework for characterizing structures representing trees for which conjunctive queries can be evaluated efficiently. Then we completely chart the tractability frontier of the problem and establish a dichotomy theorem for our axis relations, i.e., we find all subset-maximal sets of axes for which query evaluation is in polynomial time and show that for all other cases, query evaluation is NP-complete. All polynomial-time results are obtained immediately using the proof techniques from our framework. Finally, we study the expressiveness of conjunctive queries over trees and show that for each conjunctive query, there is an equivalent acyclic positive query (i.e., a set of acyclic conjunctive queries), but that in general this query is not of polynomial size.

Lexical constraints on the input of speech and on-line handwriting systems improve the performance of such systems. A significant gain in speed can be achieved by integrating in a digraph structure the different Hidden Markov Models (HMM) corresponding to the words of the relevant lexicon. This integration avoids redundant computations by sharing intermediate results between HMM's corresponding to different words of the lexicon. In this paper, we introduce a token passing method to perform simultaneously the computation of the a posteriori probabilities of all the words of the lexicon. The coding scheme that we introduce for the tokens is optimal in the information theory sense. The tokens use the minimum possible number of bits. Overall, we optimize simultaneously the execution speed and the memory requirement of the recognition systems.

The problem of designing optimal trajectory for redundant manipulators has attracted many researchers for the last three decades. One of the main reasons is the use of kinematically redundant robots is expected to increase in the future due to their increased flexibility. Some of the extra capabilities include the ability to avoid internal singularities or exte rnal obstacles over their entire workspace (Parket et al.,1989). Also, the inverse kinematics problem is underdetermined and admits an infinite number of distinct feasible solutions, meaning that a given end-effector pos es can be realized by an infinite number of distinct manipulator configurations (McAvoy, et al, 2000). In order to overcome the shortcomings inherent in non-redundant robots, redundant robots have been utilized in industrial applications to increase fl exibility and dexterity around a restricted task space in pres ence of obstacle.

This is a presentation of a new system for invariant recognition of 2D objects with overlapping classes, that can not be effectively recognized with the traditional methods. The translation, scale and partial rotation invariant contour object description is transformed in a DCT spectrum space. The obtained frequency spectrums are decomposed into frequency bands in order to feed different BPG neural nets (NNs). The NNs are structured in three stages - filtering and full rotation invariance; partial recognition; general classification. The designed multi-stage BPG Neural Structure shows very good accuracy and flexibility when tested with 2D objects used in the discontinuous production. The reached speed and the opportunuty for an easy restructuring and reprogramming of the system makes it suitable for application in different applied systems for real time work.

There is a belief that complexity and chaos are essential for adaptability. But life deals with complexity every moment, without the chaos that engineers fear so, by invoking goal-directed behaviour. Goals can be programmed. That is why living organisms give us hope to achieve adaptability in robots. In this paper a method for the description of a goal-directed, or programmed, behaviour, interacting with uncertainty of environment, is described. We suggest reducing the structural (goals, intentions) and stochastic components (probability to realise the goal) of individual behaviour to random variables with nominal values to apply probabilistic approach. This allowed us to use a Normalized Entropy Index to detect the system state by estimating the contribution of each agent to the group behaviour. The number of possible group states is 27. We argue that adaptation has a limited number of possible paths between these 27 states. Paths and states can be programmed so that after adjustment to any particular case of task and conditions, adaptability will never involve chaos. We suggest the application of the model to operation of robots or other devices in remote and/or dangerous places.

This paper is a summary of the recently accomplished research work on flexible gripping systems. The goal is to develop a gripper which can be used for a great amount of geometrically variant workpieces. The economic aspect is of particular importance during the whole development. The high flexibility of the gripper is obtained by three parallel used principles. These are human and computer based analysis of the gripping object as well as mechanical adaptation of the gripper to the object with the help of servo motors. The focus is on the gripping of free-form surfaces with suction cup.

Its capabilities cover a wide range of possible styles of algorithms. It is al so a standard environment for testing different techniques of making standard software agents with distributed architecture[10]. Rescue Simulation System also prov ides a standard framework for testing proposed algorithms and mathematical models of disaster events[8]. Designing an autonomous agent set like the one that is required for RoboCup Rescue Simulation is a little bit more of a challenge. Planning effective collaboration for a Multi-Agent team in disastrous environments still remains a challenging area in AI. Efforts of Multi-Agent researchers have provided somewhat of a standard in modeling and designing software. A lot of effort has gone into reaching coordination between different agents and making autonomous decisions that work toward the team goal[9]. But practical results in complicated domains such as RoboCup Rescue Simulation indicate that heuristic criteria still remain as a major part of a successful system[11]. This may signal lack of satisfactory models for these complicated situations.