28 December 2012 in IEEE Transactions on Signal Processing
Blind equalizers avoid the transmission of pilot sequences, allowing a more efficient use of the channel bandwidth. Normally, after a first rough equalization is achieved, it is necessary to switch these equalizers to a decision-directed (DD) mode to reduce the steady-state mean-square error (MSE) to acceptable levels. The selection of an appropriate MSE threshold for switching between the blind and the DD modes is critical to obtain a good overall performance; however, this is not an easy task, since it depends on several factors such as the signal constellation, the communication channel, or the signal-to-noise ratio. In this paper, we propose an equalization scheme that adaptively combines a blind and a DD equalizers running in parallel. The combination is itself adapted in a blind manner, and as a result the overall scheme can automatically switch between the component filters, avoiding the need to set the transition MSE level a priori. The performance of our proposal is illustrated both analytically and through an extensive set of simulations, where we show its advantages with respect to existing hard- and soft-switching equalization schemes.
30 March 2022 in Scientometrics
Probabilistic topic modeling algorithms like Latent Dirichlet Allocation (LDA) have become powerful tools for the analysis of large collections of documents (such as papers, projects, or funding applications) in science, technology an innovation (STI) policy design and monitoring. However, selecting an appropriate and stable topic model for a specific application (by adjusting the hyperparameters of the algorithm) is not a trivial problem. Common validation metrics like coherence or perplexity, which are focused on the quality of topics, are not a good fit in applications where the quality of the document similarity relations inferred from the topic model is especially relevant. Relying on graph analysis techniques, the aim of our work is to state a new methodology for the selection of hyperparameters which is specifically oriented to optimize the similarity metrics emanating from the topic model. In order to do this, we propose two graph metrics: the first measures the variability of the similarity graphs that result from different runs of the algorithm for a fixed value of the hyperparameters, while the second metric measures the alignment between the graph derived from the LDA model and another obtained using metadata available for the corresponding corpus. Through experiments on various corpora related to STI, it is shown that the proposed metrics provide relevant indicators to select the number of topics and build persistent topic models that are consistent with the metadata. Their use, which can be extended to other topic models beyond LDA, could facilitate the systematic adoption of this kind of techniques in STI policy analysis and design.