Publish in OALib Journal

ISSN: 2333-9721

APC: Only $99


Any time

2019 ( 17 )

2018 ( 10 )

2017 ( 9 )

2016 ( 9 )

Custom range...

Search Results: 1 - 10 of 884 matches for " Ernest Fokoué "
All listed articles are free for downloading (OA Articles)
Page 1 /884
Display every page Item
A Comparison of Classifiers in Performing Speaker Accent Recognition Using MFCCs  [PDF]
Zichen Ma, Ernest Fokoué
Open Journal of Statistics (OJS) , 2014, DOI: 10.4236/ojs.2014.44025

An algorithm involving Mel-Frequency Cepstral Coefficients (MFCCs) is provided to perform signal feature extraction for the task of speaker accent recognition. Then different classifiers are compared based on the MFCC feature. For each signal, the mean vector of MFCC matrix is used as an input vector for pattern recognition. A sample of 330 signals, containing 165 US voice and 165 non-US voice, is analyzed. By comparison, k-nearest neighbors yield the highest average test accuracy, after using a cross-validation of size 500, and least time being used in the computation.

Probit Normal Correlated Topic Model  [PDF]
Xingchen Yu, Ernest Fokoué
Open Journal of Statistics (OJS) , 2014, DOI: 10.4236/ojs.2014.411083
Abstract: The logistic normal distribution has recently been adapted via the transformation of multivariate Gaussian variables to model the topical distribution of documents in the presence of correlations among topics. In this paper, we propose a probit normal alternative approach to modelling correlated topical structures. Our use of the probit model in the context of topic discovery is novel, as many authors have so far concentrated solely of the logistic model partly due to the formidable inefficiency of the multinomial probit model even in the case of very small topical spaces. We herein circumvent the inefficiency of multinomial probit estimation by using an adaptation of the diagonal orthant multinomial probit in the topic models context, resulting in the ability of our topic modeling scheme to handle corpuses with a large number of latent topics. An additional and very important benefit of our method lies in the fact that unlike with the logistic normal model whose non-conjugacy leads to the need for sophisticated sampling schemes, our approach exploits the natural conjugacy inherent in the auxiliary formulation of the probit model to achieve greater simplicity. The application of our proposed scheme to a well-known Associated Press corpus not only helps discover a large number of meaningful topics but also reveals the capturing of compellingly intuitive correlations among certain topics. Besides, our proposed approach lends itself to even further scalability thanks to various existing high performance algorithms and architectures capable of handling millions of documents.
Random Subspace Learning Approach to High-Dimensional Outliers Detection  [PDF]
Bohan Liu, Ernest Fokoué
Open Journal of Statistics (OJS) , 2015, DOI: 10.4236/ojs.2015.56063

We introduce and develop a novel approach to outlier detection based on adaptation of random subspace learning. Our proposed method handles both high-dimension low-sample size and traditional low-dimensional high-sample size datasets. Essentially, we avoid the computational bottleneck of techniques like Minimum Covariance Determinant (MCD) by computing the needed determinants and associated measures in much lower dimensional subspaces. Both theoretical and computational development of our approach reveal that it is computationally more efficient than the regularized methods in high-dimensional low-sample size, and often competes favorably with existing methods as far as the percentage of correct outlier detection are concerned.

Nonnegative Matrix Factorization with Zellner Penalty  [PDF]
Matthew A. Corsetti, Ernest Fokoué
Open Journal of Statistics (OJS) , 2015, DOI: 10.4236/ojs.2015.57077

Nonnegative matrix factorization (NMF) is a relatively new unsupervised learning algorithm that decomposes a nonnegative data matrix into a parts-based, lower dimensional, linear representation of the data. NMF has applications in image processing, text mining, recommendation systems and a variety of other fields. Since its inception, the NMF algorithm has been modified and explored by numerous authors. One such modification involves the addition of auxiliary constraints to the objective function of the factorization. The purpose of these auxiliary constraints is to impose task-specific penalties or restrictions on the objective function. Though many auxiliary constraints have been studied, none have made use of data-dependent penalties. In this paper, we propose Zellner nonnegative matrix factorization (ZNMF), which uses data-dependent auxiliary constraints. We assess the facial recognition performance of the ZNMF algorithm and several other well-known constrained NMF algorithms using the Cambridge ORL database.

Prediction Error Reduction Function as a Variable Importance Score
Ernest Fokoué
Statistics , 2015,
Abstract: This paper introduces and develops a novel variable importance score function in the context of ensemble learning and demonstrates its appeal both theoretically and empirically. Our proposed score function is simple and more straightforward than its counterpart proposed in the context of random forest, and by avoiding permutations, it is by design computationally more efficient than the random forest variable importance function. Just like the random forest variable importance function, our score handles both regression and classification seamlessly. One of the distinct advantage of our proposed score is the fact that it offers a natural cut off at zero, with all the positive scores indicating importance and significance, while the negative scores are deemed indications of insignificance. An extra advantage of our proposed score lies in the fact it works very well beyond ensemble of trees and can seamlessly be used with any base learners in the random subspace learning context. Our examples, both simulated and real, demonstrate that our proposed score does compete mostly favorably with the random forest score.
Bayesian Variable Selection for Linear Regression with the $κ$-$G$ Priors
Zichen Ma,Ernest Fokoué
Statistics , 2015,
Abstract: In this article we develop a new methodology for Bayesian variable selection in multiple linear regression that is independent of the standard indicator vector method. Serving as an extension of Zellner's $g$-prior, we extend the original scalar $g$ to a diagonal matrix $G$ that controls the stability of the prior on the coefficients $\boldsymbol{\beta}$, and each of the elements $g_j$ controls the stability of its corresponding dimension. From the Metropolis-within-Gibbs sampling method, the posterior values of $G$ are sampled and those promising variables tend to have a $g_j$'s that are close to $0$. Thus the promising variables are chosen based on the posterior of $g_j$. As each of the $g_j$'s is a stabilizer of its own dimension, the $1-g_j$ values imitates the posterior inclusion probability (PIP) as in the standard methodology.
Tariffs and Total Factor Productivity: The Case of Ghanaian Manufacturing Firms  [PDF]
Charles Ackah, Ernest Ernest Aryeetey, Oliver Morrissey
Modern Economy (ME) , 2012, DOI: 10.4236/me.2012.33037
Abstract: This paper investigates the effects of trade liberalization on firm productivity in Ghana. We examine Ghanaian trade policy from 1993 to 2002, a period during which trade liberalization deepened with intermittent protection in a number of ways across industries, to investigate the effects of trade policy reforms and firm productivity. We find a strong negative impact of nominal tariffs on firm productivity, controlling for observed and unobserved firm characteristics and industry heterogeneity, a result that is robust to various alterations of the base model, including treating tariffs as endogenous and employing different estimation techniques. These results indicate that firms that are overprotected have a lower level of Total Factor Productivity than firms that are exposed to import competition. The estimated coefficients on both tariffs and its squared term confirm that higher tariffs are particularly distortionary.
BMI and Risk Factors of Underweight and Obesity in HIV Subjects in Eastern Nigeria  [PDF]
Ernest Ndukaife Anyabolu
World Journal of AIDS (WJA) , 2016, DOI: 10.4236/wja.2016.61002
Abstract: Background and Objectives: Human immunodeficiency virus infection (HIV) is a global healthcare problem. Progression of HIV infection is commonly associated with decreasing weight. In the early phases of HIV infection, factors associated with weight changes are not completely known. This study evaluated the body mass index (BMI) and its potential risk factors in drug-naive HIV subjects in Owerri, Eastern Nigeria. Methodology: This was a cross-sectional study of HIV subjects. BMI was determined. Relevant investigations were performed. Potential risk factors of BMI were analyzed at different BMI categories. Association of variables with BMI and the strength of variables to predict BMI, underweight and obesity were determined. Results: The mean BMI of the HIV subjects was 26.2 ± 5.4 kg/m2. Underweight was present in 24 (6.1%), overweight in 150 (38.4%) and obesity in 84 (21.5%) of the HIV subjects. High spot urine creatinine (SUCr), high 24-hour urine osmolality (24HUOsm), high serum cholesterol and high hemoglobin predicted BMI in HIV subjects. Low 24HUOsm predicted under weight, whereas low 24-hour urine protein (24 HUP) and high 24HUOsm predicted obesity in HIV subjects. Conclusion: The prevalence of underweight was low (6.1%), overweight high (38.4%) and obesity high (21.5%) in HIV subjects. High SUCr, high 24HUOsm, high serum cholesterol and high hemoglobin were predictors of BMI in HIV subjects. Low 24HUOsm was a predictor of underweight, while low 24HUP and high 24HUOsm were predictors of obesity in HIV subjects. Abnormalities of serum lipids, renal function, and anemia were common in HIV subjects who were underweight and in those obese. Underweight HIV subjects should be evaluated at the early stages for dyslipidemia, renal damage and anemia.
Comparative Thermometery in Paediatric Age Group: Is the Non-Touch Infrared Thermometer (NTIT) Reading Comparable to Regular Mercury-in-Glass Thermometer (MIGT) Reading?  [PDF]
Yetunde Olasinde, Moninuola Ernest, Gbenga Popoola, Kolade Ernest
Open Journal of Pediatrics (OJPed) , 2018, DOI: 10.4236/ojped.2018.84031
Abstract: Background: Accurate temperature measurement is a critical step in evaluating health or disease especially in children and immmunocompromised subjects; inaccurate measurement may lead to improper diagnosis, wrong treatment or inappropriate intervention. Several methods of temperature measurements exist and comparing these gives room for choosing a near ideal method in terms of speed, safety and accuracy. The study aimed to compare the forehead non touch infra-red thermometer with the axilllary mercury-in-glass method of temperature measurement in the Paediatric age-group. Methods: Study was given ethical approval as part of a larger study. Four hundred and thirty seven children aged 1 to 24 months were studied at the well-baby/immunizationclinic of the University of Ilorin Teaching Hospital over a 6-months period. Both non-touch infrared and theregular mercury-in-glass thermometers were used to take the body temperatures. Data were analysed with SPSS version 21. Pearson correlation was used to determine the relationship between the two methods of temperature measurements, while Bland-Altman method was used to test for level of agreement between them. Results: The
Nutrient Content and Organoleptic Quality of Traditional African Strain and Rhode Island Chickens and the Effect of Feed Rations
Germain Kansci, Eric Bogne Lele, Minette, Martin Fotso, Elie Fokou
Journal of Food Technology in Africa , 2004,
Abstract: Pectoral and thigh muscles of African strain and Rhode Island chickens were characterised for their contents in moisture, proteins, lipids and phospholipids. Water retention capacity of the muscles was measured and the influence of enriched cotton cake feed on the muscle quality of Rhode Island race chicken was evaluated. The chicken stocks and their muscles were similar in terms of their water contents (73.6 - 74.9 g/100g). The muscles of the both chicken stocks showed excellent water retention capacity (up to 69%). The Rhode Island race were however, richer in lipids (2.74 - 3.46 g/100g) and phospholipids (0.34 - 0.57 g/100g). Feeding with cotton cake-enriched rations increased the lipid content further. On the other hand, muscles of the African strain chicken were richer in proteins (22.5 - 24.3 g/100g) than those of the Rhode Island race (19.5 - 22.5 g/ 100g). The nutritional, organoleptic and technological qualities of the chicken muscles are discussed with respect to these characteristics. Cultural consumption habits could explain preference of African Strain Chicken muscles by the local population. Key Words: chicken, feed, lipids, water retention capacity, proteins, quality Journal of Food Technology in Africa Vol.9(1) 2004: 26-28
Page 1 /884
Display every page Item

Copyright © 2008-2017 Open Access Library. All rights reserved.