Apprenticeship learning (AL) is a class of “learning from demonstrations” techniques where the reward function of a Markov Decision Process (MDP) is unknown to the learning agent and the agent has to derive a good policy by observing an expert’s demonstrations. In this paper, we study the problem of how to make AL algorithms inherently safe while still meeting its learning objective. We conside...

In this paper the abstraction-refinement paradigm based on 3-valued logics is extended to the setting of probabilistic systems. We define a notion of abstraction for Markov chains. To be able to relate the behavior of abstract and concrete systems, we equip the notion of abstraction with the concept of simulation. Furthermore, we present model checking for abstract probabilistic systems (abstra...

In this paper, the author aims to establish a mathematical model for a mimic computer. To this end, a novel automaton is proposed. First, a one-dimensional cellular automaton is used for expressing some dynamic changes in the structure of a computing unit, a sequential automaton is employed to describe some state transitions, a hierarchical automaton is employed to express the different granula...

This paper presents several symbolic counterexample generation algorithms for discrete-time Markov chains (DTMCs) violating a PCTL formula. A counterexample is (a symbolic representation of) a sub-DTMC that is incrementally generated. The crux to this incremental approach is the symbolic generation of paths that belong to the counterexample. We consider two approaches. First, we extend bounded ...

This paper provides a glimpse of basic probabilistic database concepts, which is an active area of research in today’s world. The discussion starts with the need for probabilistic databases, and their advantages over conventional databases in certain circumstances. Then, some of the key aspects of probabilistic databases are discussed, which include topics like types of uncertainties in a proba...

We replace deterministic block encryption by probabilistic encryption of single. This property enabled Goldwasser and Micali ill to device a scheme for Mental.A new probabilistic model of data encryption is introduced. Laboratory of.presenting the first probabilistic cryptosystem by Goldwasser and Micali.

One of the key challenges in human action recognition from video sequences is how to model an action sufficiently. Therefore, in this paper we propose a novel motion-based representation called Motion Context (MC), which is insensitive to the scale and direction of an action, by employing image representation techniques. A MC captures the distribution of the motion words (MWs) over relative loc...

Building robust stochastic language models is a major issue in speech recognition systems. Conventional word-based n-gram models do not capture any linguistic constraints inherent in speech. In this paper the notion of function and content words (open/closed word classes) is used to provide linguistic knowledge that can be incorporated into language models. Function words are articles, preposit...

A human annotator can provide hints to a machine learner by highlighting contextual “rationales” for each of his or her annotations (Zaidan et al., 2007). How can one exploit this side information to better learn the desired parameters θ? We present a generative model of how a given annotator, knowing the true θ, stochastically chooses rationales. Thus, observing the rationales helps us infer t...

The paper describes the results of an empirical study of integrating bigram collocations and similarities between them and unigrams into topic models. First of all, we propose a novel algorithm PLSA-SIM that is a modification of the original algorithm PLSA. It incorporates bigrams and maintains relationships between unigrams and bigrams based on their component structure. Then we analyze a vari...