PERFORMANCE OF DIFFERENT CLASSIFIERS IN SPEECH RECOGNITION

Speech is the most natural means of communication among human beings and speech processing and recognition are intensive areas of research for the last five decades. Since speech recognition is a pattern recognition problem, classification is an important part of any speech recognition system. In this work, a speech recognition system is developed for recognizing speaker independent spoken digits in Malayalam. Voice signals are sampled directly from the microphone. The proposed method is implemented for 1000 speakers uttering 10 digits each. Since the speech signals are affected by background noise, the signals are tuned by removing the noise from it using wavelet denoising method based on Soft Thresholding. Here, the features from the signals are extracted using Discrete Wavelet Transforms (DWT) because they are well suitable for processing non-stationary signals like speech. This is due to their multi- resolutional, multi-scale analysis characteristics. Speech recognition is a multiclass classification problem. So, the feature vector set obtained are classified using three classifiers namely, Artificial Neural Networks (ANN), Support Vector Machines (SVM) and Naive Bayes classifiers which are capable of handling multiclasses. During classification stage, the input feature vector data is trained using information relating to known patterns and then they are tested using the test data set. The performances of all these classifiers are evaluated based on recognition accuracy. All the three methods produced good recognition accuracy. DWT and ANN produced a recognition accuracy of 89%, SVM and DWT combination produced an accuracy of 86.6% and Naive Bayes and DWT combination produced an accuracy of 83.5%. ANN is found to be better among the three methods.

Mechanical Properties of Steel Fiber-Reinforced Concrete

This paper presents the results from an experimental program and an analytical assessment of the influence of addition of fibers on mechanical properties of concrete. Models derived based on the regression analysis of 60 test data for various mechanical properties of steel fiber-reinforced concrete have been presented. The various strength properties studied are cube and cylinder compressive strength, split tensile strength, modulus of rupture and postcracking performance, modulus of elasticity, Poisson’s ratio, and strain corresponding to peak compressive stress. The variables considered are grade of concrete, namely, normal strength 35 MPa , moderately high strength 65 MPa , and high-strength concrete 85 MPa , and the volume fraction of the fiber Vf =0.0, 0.5, 1.0, and 1.5% . The strength of steel fiber-reinforced concrete predicted using the proposed models have been compared with the test data from the present study and with various other test data reported in the literature. The proposed model predicted the test data quite accurately. The study indicates that the fiber matrix interaction contributes significantly to enhancement of mechanical properties caused by the introduction of fibers, which is at variance with both existing models and formulations based on the law of mixtures

Indicators of nitrate in wetland surface and soil-waters: interactions of vegetation and environmental factors

This paper describes a new bio-indicator method for assessing wetland ecosystem health: as such, the study is particularly relevant to current legislation such as the EU Water Framework Directive, which provides a baseline of the current status Of Surface waters. Seven wetland sites were monitored across northern Britain, with model construction data for predicting, eco-hydroloplical relationships collected from five sites during 1999, Two new sites and one repeat site were monitored during 2000 to provide model test data. The main growing season for the vegetation, and hence the sampling period, was May-August during both years. Seasonal mean concentrations of nitrate (NO3-) in surface and soil water samples during 1999 ranged from 0.01 to 14.07 mg N 1(-1), with a mean value of 1.01 mg N 1(-1). During 2000, concentrations ranged from trace level (<0.01 m- N 1(-1)) to 9.43 mg N 1(-1), with a mean of 2.73 mg N 1(.)(-1) Surface and soil-water nitrate concentrations did not influence plant species composition significantly across representative tall herb fen and mire communities. Predictive relationships were found between nitrate concentrations and structural characteristics of the wetland vegetation, and a model was developed which predicted nitrate concentrations from measures of plant diversity, canopy structure and density of reproductive structures. Two further models, which predicted stem density and density of reproductive structures respectively, utilised nitrate concentration as one of the independent predictor variables. Where appropriate, the models were tested using data collected during 2000. This approach is complementary to species-based monitoring, representing a useful and simple too] to assess ecological status in target wetland systems and has potential for bio-indication purposes.

Constrained models of evolution lead to improved prediction of functional linkage from correlated gain and loss of genes

Motivation: We compare phylogenetic approaches for inferring functional gene links. The approaches detect independent instances of the correlated gain and loss of pairs of genes from species' genomes. We investigate the effect on results of basing evidence of correlations on two phylogenetic approaches, Dollo parsminony and maximum likelihood (ML). We further examine the effect of constraining the ML model by fixing the rate of gene gain at a low value, rather than estimating it from the data. Results: We detect correlated evolution among a test set of pairs of yeast (Saccharomyces cerevisiae) genes, with a case study of 21 eukaryotic genomes and test data derived from known yeast protein complexes. If the rate at which genes are gained is constrained to be low, ML achieves by far the best results at detecting known functional links. The model then has fewer parameters but it is more realistic by preventing genes from being gained more than once. Availability: BayesTraits by M. Pagel and A. Meade, and a script to configure and repeatedly launch it by D. Barker and M. Pagel, are available at http://www.evolution.reading.ac.uk .

Dual-orthogonal radial basis function networks for nonlinear time series prediction

A new structure of Radial Basis Function (RBF) neural network called the Dual-orthogonal RBF Network (DRBF) is introduced for nonlinear time series prediction. The hidden nodes of a conventional RBF network compare the Euclidean distance between the network input vector and the centres, and the node responses are radially symmetrical. But in time series prediction where the system input vectors are lagged system outputs, which are usually highly correlated, the Euclidean distance measure may not be appropriate. The DRBF network modifies the distance metric by introducing a classification function which is based on the estimation data set. Training the DRBF networks consists of two stages. Learning the classification related basis functions and the important input nodes, followed by selecting the regressors and learning the weights of the hidden nodes. In both cases, a forward Orthogonal Least Squares (OLS) selection procedure is applied, initially to select the important input nodes and then to select the important centres. Simulation results of single-step and multi-step ahead predictions over a test data set are included to demonstrate the effectiveness of the new approach.

Superensembles of linear viscoelastic models of polymer melts

The idea of incorporating multiple models of linear rheology into a superensemble, to forge a consensus forecast from the individual model predictions, is investigated. The relative importance of the individual models in the so-called multimodel superensemble (MMSE) was inferred by evaluating their performance on a set of experimental training data, via nonlinear regression. The predictive ability of the MMSE model was tested by comparing its predictions on test data that were similar (in-sample) and dissimilar (out-of-sample) to the training data used in the calibration. For the in-sample forecasts, we found that the MMSE model easily outperformed the best constituent model. The presence of good individual models greatly enhanced the MMSE forecast, while the presence of some bad models in the superensemble also improved the MMSE forecast modestly. While the performance of the MMSE model on the out-of-sample training data was not as spectacular, it demonstrated the robustness of this approach.

Random Prism: a noise-tolerant alternative to Random Forests

Ensemble learning can be used to increase the overall classification accuracy of a classifier by generating multiple base classifiers and combining their classification results. A frequently used family of base classifiers for ensemble learning are decision trees. However, alternative approaches can potentially be used, such as the Prism family of algorithms that also induces classification rules. Compared with decision trees, Prism algorithms generate modular classification rules that cannot necessarily be represented in the form of a decision tree. Prism algorithms produce a similar classification accuracy compared with decision trees. However, in some cases, for example, if there is noise in the training and test data, Prism algorithms can outperform decision trees by achieving a higher classification accuracy. However, Prism still tends to overfit on noisy data; hence, ensemble learners have been adopted in this work to reduce the overfitting. This paper describes the development of an ensemble learner using a member of the Prism family as the base classifier to reduce the overfitting of Prism algorithms on noisy datasets. The developed ensemble classifier is compared with a stand-alone Prism classifier in terms of classification accuracy and resistance to noise.

User-friendly free-text keystroke dynamics authentication for practical applications

This paper introduces a novel approach for free-text keystroke dynamics authentication which incorporates the use of the keyboard’s key-layout. The method extracts timing features from specific key-pairs. The Euclidean distance is then utilized to find the level of similarity between a user’s profile data and his/her test data. The results obtained from this method are reasonable for free-text authentication while maintaining the maximum level of user relaxation. Moreover, it has been proven in this study that flight time yields better authentication results when compared with dwell time. In particular, the results were obtained with only one training sample for the purpose of practicality and ease of real life application.

Cloud clearing techniques over land for land surface temperature retrieval from the Advanced Along Track Scanning Radiometer

We present five new cloud detection algorithms over land based on dynamic threshold or Bayesian techniques, applicable to the Advanced Along Track Scanning Radiometer (AATSR) instrument and compare these with the standard threshold based SADIST cloud detection scheme. We use a manually classified dataset as a reference to assess algorithm performance and quantify the impact of each cloud detection scheme on land surface temperature (LST) retrieval. The use of probabilistic Bayesian cloud detection methods improves algorithm true skill scores by 8-9 % over SADIST (maximum score of 77.93 % compared to 69.27 %). We present an assessment of the impact of imperfect cloud masking, in relation to the reference cloud mask, on the retrieved AATSR LST imposing a 2 K tolerance over a 3x3 pixel domain. We find an increase of 5-7 % in the observations falling within this tolerance when using Bayesian methods (maximum of 92.02 % compared to 85.69 %). We also demonstrate that the use of dynamic thresholds in the tests employed by SADIST can significantly improve performance, applicable to cloud-test data to provided by the Sea and Land Surface Temperature Radiometer (SLSTR) due to be launched on the Sentinel 3 mission (estimated 2014).

Are sulfurous soil amendments (S0, Fe(II)SO4, Fe(III)SO4) an effective tool in the restoration of heathland and acidic grassland after four decades of rock phosphate Fertilization?

This paper deals with the complex issue of reversing long-term improvements of fertility in soils derived from heathlands and acidic grasslands using sulfur-based amendments. The experiment was conducted on a former heathland and acid grassland in the U.K. that was heavily fertilized and limed with rock phosphate, chalk, and marl. The experimental work had three aims. First, to determine whether sulfurous soil amendments are able to lower pH to a level suitable for heathland and acidic grassland re-creation (approximately 3 pH units). Second, to determine what effect the soil amendments have on the available pool of some basic cations and some potentially toxic acidic cations that may affect the plant community. Third, to determine whether the addition of Fe to the soil system would sequester PO4− ions that might be liberated from rock phosphate by the experimental treatments. The application of S0 and Fe(II)SO4− to the soil was able to reduce pH. However, only the highest S0 treatment (2,000 kg/ha S) lowered pH sufficiently for heathland restoration purposes but effectively so. Where pH was lowered, basic cations were lost from the exchangeable pool and replaced by acidic cations. Where Fe was added to the soil, there was no evidence of PO4− sequestration from soil test data (Olsen P), but sequestration was apparent because of lower foliar P in the grass sward. The ability of the forb Rumex acetosella to apparently detoxify Al3+, prevalent in acidified soils, appeared to give it a competitive advantage over other less tolerant species. We would anticipate further changes in plant community structure through time, driven by Al3+ toxicity, leading to the competitive exclusion of less tolerant species. This, we suggest, is a key abiotic driver in the restoration of biotic (acidic plant) communities.

A Genetic Algorithm for the selection of structural MRI features for classification of Mild Cognitive Impairment and Alzheimer's Disease

This work investigates the problem of feature selection in neuroimaging features from structural MRI brain images for the classification of subjects as healthy controls, suffering from Mild Cognitive Impairment or Alzheimer’s Disease. A Genetic Algorithm wrapper method for feature selection is adopted in conjunction with a Support Vector Machine classifier. In very large feature sets, feature selection is found to be redundant as the accuracy is often worsened when compared to an Support Vector Machine with no feature selection. However, when just the hippocampal subfields are used, feature selection shows a significant improvement of the classification accuracy. Three-class Support Vector Machines and two-class Support Vector Machines combined with weighted voting are also compared with the former and found more useful. The highest accuracy achieved at classifying the test data was 65.5% using a genetic algorithm for feature selection with a three-class Support Vector Machine classifier.

Bayesian analysis of skew-t multivariate null intercept measurement error model

The multivariate skew-t distribution (J Multivar Anal 79:93-113, 2001; J R Stat Soc, Ser B 65:367-389, 2003; Statistics 37:359-363, 2003) includes the Student t, skew-Cauchy and Cauchy distributions as special cases and the normal and skew-normal ones as limiting cases. In this paper, we explore the use of Markov Chain Monte Carlo (MCMC) methods to develop a Bayesian analysis of repeated measures, pretest/post-test data, under multivariate null intercept measurement error model (J Biopharm Stat 13(4):763-771, 2003) where the random errors and the unobserved value of the covariate (latent variable) follows a Student t and skew-t distribution, respectively. The results and methods are numerically illustrated with an example in the field of dentistry.

Photovoltaic Process Development and innovative Techniques

Photovoltaic processing is one of the processes that have significance in semiconductor process line. It is complicated due to the no. of elements involved that directly or indirectly affect the processing and final yield. So mathematically or empirically we can’t say assertively about the results specially related with diffusion, antireflective coating and impurity poisoning. Here I have experimented and collected data on the mono-crystal silicon wafers with varying properties and outputs. Then by using neural network with available experimental data output required can be estimated which is further tested by the test data for authenticity. One can say that it’s a kind of process simulation with varying input of raw wafers to get desired yield of photovoltaic mono-crystal cells.

Skicklig läs- och skrivundervisning i åk 1-3 : Om lärares möjligheter och hinder

This study focuses on teachers’ opportunities and obstacles to perform skillful reading and writing instruction. It’s about the ability to accurately identify where students are in their reading and writing process and to help them develop good reading skills. It is also about the ability to recognize signs of difficulties that students may have in their written language development and to know what efforts are needed to help them advance their reading and writing skills. The research is based on teachers’ own statements and survey responses on the external conditions for teaching and on their approach, attitudes and knowledge in reading and writing. The empirical material consists of interviews, surveys and test data. The interview study was conducted with eight teachers. The questionnaire was answered by 249 teachers, while the knowledge test was conducted of 269 teachers and 31 special education teachers. Many of the teachers in this study have lack knowledge in the structure of language and common Swedish spelling rules. Furthermore, it appears that a large part of them are unaccustomed to explaining, in detail, students’ reading development and find it difficult to systematically describe the aspects of daily literacy instruction. The overall picture is that many teachers teach without having tools to reflect on how their education really affects students’ reading and writing. These shortcomings make it difficult to conduct effective literacy instruction. Once students have learned to decode or if they have reading difficulties, many teachers seem to one-sidedly focus on getting students to read more. The consequence could be that those who would need to practice more on the technical basic of reading or comprehension strategies are left without support. Lack of variety and individuality in fluency and comprehension training can challenge the students’ reading and writing development. The teachers in the study, who have the old junior school teacher and elementary teacher education, have the highest amount of knowledge of reading and writing (the test). Good education can provide student teachers with professional skills that they may develop further in their careers. Knowledge of the meaning of phonological and phonemic awareness as well as knowledge of how to count phonemes seem to be important for knowledge of reading and writing (the test). Knowledge of basic reading processes can be obtained by systematic and structured work with students’ linguistic development, and through continuous dialogues with experienced colleagues on how and why questions. This is one important way to work also in teacher training. When essential professional skills are established in the teacher education, in practice students will obtain the school’s learning goals.

A simplified heat pump model for use in solar plus heat pump system simulation studies

Solar plus heat pump systems are often very complex in design, with sometimes special heat pump arrangements and control. Therefore detailed heat pump models can give very slow system simulations and still not so accurate results compared to real heat pump performance in a system. The idea here is to start from a standard measured performance map of test points for a heat pump according to EN 14825 and then determine characteristic parameters for a simplified correlation based model of the heat pump. By plotting heat pump test data in different ways including power input and output form and not only as COP, a simplified relation could be seen. By using the same methodology as in the EN 12975 QDT part in the collector test standard it could be shown that a very simple model could describe the heat pump test data very accurately, by identifying 4 parameters in the correlation equation found. © 2012 The Authors.