Asymptotic behavior of a hierarchical system of learning automata

Learning automata arranged in a two-level hierarchy are considered. The automata operate in a stationary random environment and update their action probabilities according to the linear-reward- -penalty algorithm at each level. Unlike some hierarchical systems previously proposed, no information transfer exists from one level to another, and yet the hierarchy possesses good convergence properties. Using weak-convergence concepts it is shown that for large time and small values of parameters in the algorithm, the evolution of the optimal path probability can be represented by a diffusion whose parameters can be computed explicitly.

A hierarchical system of learning automata that can learn die globally optimal path

Systems of learning automata have been studied by various researchers to evolve useful strategies for decision making under uncertainity. Considered in this paper are a class of hierarchical systems of learning automata where the system gets responses from its environment at each level of the hierarchy. A classification of such sequential learning tasks based on the complexity of the learning problem is presented. It is shown that none of the existing algorithms can perform in the most general type of hierarchical problem. An algorithm for learning the globally optimal path in this general setting is presented, and its convergence is established. This algorithm needs information transfer from the lower levels to the higher levels. Using the methodology of estimator algorithms, this model can be generalized to accommodate other kinds of hierarchical learning tasks.

A Hierarchical System of Learning Automata

A learning automaton operating in a random environment updates its action probabilities on the basis of the reactions of the environment, so that asymptotically it chooses the optimal action. When the number of actions is large the automaton becomes slow because there are too many updatings to be made at each instant. A hierarchical system of such automata with assured c-optimality is suggested to overcome that problem.The learning algorithm for the hierarchical system turns out to be a simple modification of the absolutely expedient algorithm known in the literature. The parameters of the algorithm at each level in the hierarchy depend only on the parameters and the action probabilities of the previous level. It follows that to minimize the number of updatings per cycle each automaton in the hierarchy need have only two or three actions.

A cooperative game of a pair of learning automata

A cooperative game played in a sequential manner by a pair of learning automata is investigated in this paper. The automata operate in an unknown random environment which gives a common pay-off to the automata. Necessary and sufficient conditions on the functions in the reinforcement scheme are given for absolute monotonicity which enables the expected pay-off to be monotonically increasing in any arbitrary environment. As each participating automaton operates with no information regarding the other partner, the results of the paper are relevant to decentralized control.

Functional transfer of Salmonella pathogenicity island 2 to Salmonella bongori and Escherichia coli.

The type III secretion system (T3SS) encoded by the Salmonella pathogenicity island 2 (SPI2) has a central role in systemic infections by Salmonella enterica and for the intracellular phenotype. Intracellular S. enterica uses the SPI2-encoded T3SS to translocate a set of effector proteins into the host cell, which modify host cell functions, enabling intracellular survival and replication of the bacteria. We sought to determine whether specific functions of the SPI2-encoded T3SS can be transferred to heterologous hosts Salmonella bongori and Escherichia coli Mutaflor, species that lack the SPI2 locus and loci encoding effector proteins. The SPI2 virulence locus was cloned and functionally expressed in S. bongori and E. coli. Here, we demonstrate that S. bongori harboring the SPI2 locus is capable of secretion of SPI2 substrate proteins under culture conditions, as well as of translocation of effector proteins under intracellular conditions. An SPI2-mediated cellular phenotype was induced by S. bongori harboring the SPI2 if the sifA locus was cotransferred. An interference with the host cell microtubule cytoskeleton, a novel SPI2-dependent phenotype, was observed in epithelial cells infected with S. bongori harboring SPI2 without additional effector genes. S. bongori harboring SPI2 showed increased intracellular persistence in a cell culture model, but SPI2 transfer was not sufficient to confer to S. bongori systemic pathogenicity in a murine model. Transfer of SPI2 to heterologous hosts offers a new tool for the study of SPI2 functions and the phenotypes of individual effectors.

Transfer of saline tolerance from one strain of rice to another by injection of DNA

Direct injection of genomic DNA from salt tolerant cv. Pokkali into developing floral tillers on IR20 produced transgenic seeds similar to Pokkali in husk colour and which germinated well in 0.2 M NaCl and had a 4-6-fold higher proline content.

Plasmid DNA mediated transfer of the herpes simplex virus thymidine kinase gene to a new bromodeoxyuridine resistant variant of human primary lung carcinoma cells

A human primary lung carcinoma cell line (HPL-R1) established from the tumor biopsy of a lung cancer patient, lacking in cytochrome P1-450 [aryl hydrocarbon (benzo[a]pyrene) hydroxylase (AHH)], was cloned and used to obtain variants deficient in the expression of thymidine-kinase via treatment with 5-bromo-2'-deoxyuridine, and selection for drug resistance phenotype. The variant cell line, precharacterized for thymidine kinase negative phenotype, was transfected with the thymidine kinase gene bearing p R-tk and px1-tk plasmids. Transfections from both the plasmids, demonstrated a frequency of 5.5 X 10(-5). The transfectants showed a 76-100% retention of the transferred phenotype. These data suggest that transfection in variant human cells can approach significant levels of stability observed with rodent cell recipients.

Transfer of F- in the reaction of SF6- with SOF4: Implications for SOF4 production in corona discharges

The temperature (T) and electric field-to-gas pressure (E/P) dependences of the rate coefficientk for the reaction SF 6 � +SOF4rarrSOF 5 � +SF5 have been measured. ForT<270>k approaches a constant of 2.1×10�9 cm3/s, and for 433>T>270 K,k decreases withT according tok (cm3/s)=0.124 exp [�3.3 lnT(K)]. ForE/Pk has a constant value of about 2.5×10�10 cm3/s, and for 130 V/cm·torr>E/P>60 V/cm·torr, the rate is approximately given byk (cm3/s)sim7.0×10�10 exp (�0.022E/P). The measured rate coefficient is used to estimate the influence of this reaction on SOF4 production from negative, point-plane, glow-type corona discharges in gas mixtures containing SF6 and at least trace amounts of O2 and H2O. A chemical kinetics model of the ion-drift region in the discharge gap is used to fit experimental data on SOF4 yields assuming that the SF 6 � +SOF4 reaction is the predominant SOF4 loss mechanism. It is found that the contribution of this reaction to SOF4 destruction falls considerably below the estimated maximum effect assuming that SF 6 � is the predominant charge carrier which reacts only with SOF4. The results of this analysis suggest that SF 6 � is efficiently deactivated by other reactions, and the influence of SF 6 � +SOF4 on SOF4 production is not necessarily more significant than that of other slower secondary processes such as gas-phase hydrolysis

Interfamilial transfer of amber suppressor gene for the isolation of amber mutants of mycobacteriophage I3

A suppressor-containing strain of Mycobacterium smegmatis SN2 was isolated by transferring an amber suppressor carried on the plasmid of Pseudomonas pseudoalcaligenes ERA through transformation. Amber mutants of mycobacteriophage I3 were isolated.

Cross-Guided Clustering: Transfer of Relevant Supervision across Tasks

Lack of supervision in clustering algorithms often leads to clusters that are not useful or interesting to human reviewers. We investigate if supervision can be automatically transferred for clustering a target task, by providing a relevant supervised partitioning of a dataset from a different source task. The target clustering is made more meaningful for the human user by trading-off intrinsic clustering goodness on the target task for alignment with relevant supervised partitions in the source task, wherever possible. We propose a cross-guided clustering algorithm that builds on traditional k-means by aligning the target clusters with source partitions. The alignment process makes use of a cross-task similarity measure that discovers hidden relationships across tasks. When the source and target tasks correspond to different domains with potentially different vocabularies, we propose a projection approach using pivot vocabularies for the cross-domain similarity measure. Using multiple real-world and synthetic datasets, we show that our approach improves clustering accuracy significantly over traditional k-means and state-of-the-art semi-supervised clustering baselines, over a wide range of data characteristics and parameter settings.