We propose an extension of the Viterbi algorithm that makes second-order hidden Markov models computationally efficient. A comparative study between first-order (HMM1's) and second-order Markov models (HMM2's) is carried out. Experimental results show that HMM2's provide a better state occupancy modeling and, alone, have performances comparable with HMM1's plus postprocessing.

No abstract

No abstract

In agricultural landscapes, methods to identify and describe meaningful landscape patterns play an important role to understand the interaction between landscape organization and ecological processes. We propose an innovative stochastic modelling method of agricultural landscape organization where the temporal regularities in land-use are first identified through recognized Land-Use Successions (LUS) before locating these successions in landscapes. These time-space regularities within landscapes are extracted using a new data mining method based on Hidden Markov Models. We applied this methodological proposal to the Niort Plain (West of France). We built a temporo-spatial analysis for this case study through spatially explicit analysis of Land Use Succession (LUS) dynamics. Implications and perspectives of such an approach, which links together the temporal and the spatial dimensions of the agricultural organization, are discussed by assessing the relationship between the agricultural landscape patterns defined using this approach and ecological data through an illustrative example of bird nests.

Summary 1.It is well established that pollen-mediated gene flow among natural plant populations depends on a complex interaction between the spatial distribution of pollen sources and the short-and long-distance components of pollen dispersal. Despite this knowledge, spatial isolation strategies proposed in Europe to ensure the harvest purity of conventional crops are based on distance from the nearest genetically modified (GM) crop and on empirical data from two-plot experiments. Here, we investigate the circumstances under which the multiplicity of pollen sources over the landscape should be considered in strategies to contain GM crops. 2. We simulated pollen dispersal over eighty 6 × 6 km simulated landscapes differing in field characteristics and in amount of GM and conventional maize. Pollen dispersal was modelled either via a Normal Inverse Gaussian (NIG, currently used for European coexistence studies) or a bivariate Student (2Dt) kernel. These kernels differ in their amount of short-and long-distance dispersal. We used linear models to analyse the impact of local and landscape variables on impurity rates (i.e. proportion of seeds sired by pollen from a transgenic crop) in conventional fields and quantified their increase due to dispersal from other than the closest GM crops. 3. The average impurity rate over a landscape increased linearly with the proportion of GM maize over that landscape. The increase was twice as fast using the NIG kernel and was governed by the short-distance dispersal component. 4. Variation in impurity rates largely depended on the distance to the closest GM crop and the size of the receptor field. However, impurity rates were generally underestimated when only dispersal from the closest GM field was considered. 5. Synthesis and applications. Distance to the closest GM crop had most impact on impurity rates in conventional fields. However, impurity rates also depended on intermediate-to long-distance dispersal from distant GM crops. Therefore, isolation distances as currently defined will probably not allow long-term coexistence of GM and conventional crops, especially as the proportion of GM crops grown increases. We suggest strategies to account for this impact of long-distance dispersal.

No abstract

In the field of speech recognition by stochastic methods, it is conventional to pursue approaches using firstorder Hidden Markov Models (HMMls). Despite the success of this approach, it is still worth investigating if some of the drawbacks of HMMls can be overcome, e.g. by using higher-order Markov processes. In this paper, we show that second-order Hidden Markov Models (HMM2s) can yield high performances in the context of continuous speech recognition. We first present the underlying equations and complexity of HMM2s in the Maximum Likelihood Estimation (MLE) paradigm. Then, we show that in a connected word recognition task, such as spelled name recognition over the telephone, HMM2s outperform HMMls. In the field of phoneme-based continuous speech recognition, we show that context-independent HMM2s can achieve more than 69% phone accuracy. 0-7803-3 192-3/96 $5.0001996 IEEE

A combination of gene loss and acquisition through horizontal gene transfer (HGT) is thought to drive Streptococcus thermophilus adaptation to its niche, i.e. milk. In this study, we describe an in silico analysis combining a stochastic data mining method, analysis of homologous gene distribution and the identification of features frequently associated with horizontally transferred genes to assess the proportion of the S. thermophilus genome that could originate from HGT. Our mining approach pointed out that about 17.7% of S. thermophilus genes (362 CDSs of 1,915) showed a composition bias; these genes were called 'atypical'. For 22% of them, their functional annotation strongly support their acquisition through HGT and consisted mainly in genes encoding mobile genetic recombinases, exopolysaccharide (EPS) biosynthesis enzymes or resistance mechanisms to bacteriophages. The distribution of the atypical genes in the Firmicutes phylum as well as in S. thermophilus species was sporadic and supported the HGT prediction for more than a half (52%, 189). Among them, 46 were found specific to S. thermophilus. Finally, by combining our method, gene annotation and sequence specific features, new genome islands were suggested in the S. thermophilus genome.