First provide a summary of the paper, and then address the following criteria: Quality, clarity, originality and significance. The paper presents a method for learning layers of representation and for completing missing queries both in input and labels in single procedure unlike some other methods like deep boltzmann machines (DBM). It is a recurrent net following the same operations as DBM with the goal of predicting a subset of inputs from its complement. Parts of paper are badly written, especially model explanation and multi-inference section, nevertheless the paper should be published and I hope the authors will rewrite them. Details: - The procedure is taken from DBM, however other then that, is there a relation between the DBM and this algorithm, or should we just treat the algorithm as one particular function (recurrent net (RNN)) that predicts subset of inputs from its complement?

Probabilistic graphical models are a key tool in machine learning applications. Computing the partition function, i.e., normalizing constant, is a fundamental task of statistical inference but it is generally computationally intractable, leading to extensive study of approximation methods. Iterative variational methods are a popular and successful family of approaches. However, even state of the art variational methods can return poor results or fail to converge on difficult instances. In this paper, we instead consider computing the partition function via sequential summation over variables. We develop robust approximate algorithms by combining ideas from mini-bucket elimination with tensor network and renormalization group methods from statistical physics. The resulting "convergence-free" methods show good empirical performance on both synthetic and real-world benchmark models, even for difficult instances.

Individuals and groups, within organisations, cooperate by producing, manipulating and organizing knowledge, and by building and refining new collective knowledge. Organisations increasingly see their intellectual capital as strategic resources that must be managed effectively to achieve competitive advantage. This capital consists of the knowledge held in the minds of its members, embodied in its procedures and decision making processes, and stored in its repositories. Subsequently, it should be useful for KM systems and Collaboration systems to integrate both kinds of capabilities into a single collaborative-and-knowledge based system to support joint efforts towards a goal [1]. Decision making is one of the critical processes where we need both knowledge management (that focuses on creation, storage, sharing and use of knowledge) and collaboration (that focuses on cooperation, communication, coordination and coproduction) to make that more effective and efficient. This paper aims to explicit step-by-step the multimodal decision making (MDM) process at three levels (individual, collective and hybrid) and is organized as follows; we start with a brief overview of the literature on collaborative knowledge management. In section three, we propose formal description of MDM process. Finally, section four presents our model of MDM process basing on the proposed formal description and UML-G profile.

Machine Intelligence 5. B. Meltzer & D. Michie (eds.), 78-98. Edinburgh: Edinburgh University Press.

In Meltzer, Bernard and Donald Michie (Eds.). Machine Intelligence 5, pp. 78-98. Edinburgh University Press.