Particle competition and cooperation for semi-supervised learning with label noise

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Supervised variational relevance learning, an analytic geometric feature selection with applications to omic datasets

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Semi-supervised learning for relevance feedback on image retrieval tasks

Relevance feedback approaches have been established as an important tool for interactive search, enabling users to express their needs. However, in view of the growth of multimedia collections available, the user efforts required by these methods tend to increase as well, demanding approaches for reducing the need of user interactions. In this context, this paper proposes a semi-supervised learning algorithm for relevance feedback to be used in image retrieval tasks. The proposed semi-supervised algorithm aims at using both supervised and unsupervised approaches simultaneously. While a supervised step is performed using the information collected from the user feedback, an unsupervised step exploits the intrinsic dataset structure, which is represented in terms of ranked lists of images. Several experiments were conducted for different image retrieval tasks involving shape, color, and texture descriptors and different datasets. The proposed approach was also evaluated on multimodal retrieval tasks, considering visual and textual descriptors. Experimental results demonstrate the effectiveness of the proposed approach.

Efeito de diferentes gomas de mascar sobre o pH salivar de crianças

Objective: To evaluate the effect of different chewing gum brands on the salivary pH of children with primary dentition. Method: Forty children were selected and assigned to four groups: control (no chewing gum); sugarless chewing gum; chewing gum with casein phosphopeptide-amorphous calcium phosphate; and chewing gum with xylitol. The first saliva collection was made after supervised tooth brushing for stabilization of the oral pH. Next, all children were instructed to drink slowly 100 mL of a cola-based soft drink (Coca-Cola®) and a new saliva collection was made 10 min later. Then, each group chewed on the chewing gum for 5 min and discarded it after this time. Saliva was collected again at 5, 10 and 15 min intervals after start using the chewing gum. Measurement of salivary pH was made with colorimetric test papers and a digital pH-meter. Data were analyzed statistically by analysis of variance and Tukey’s test at a 5% significance level. Results: The use of chewing gums accelerated the increase of salivary pH to considerably alkaline levels after consumption of an acidic beverage, especially within the first minutes. The highest levels were obtained in the groups of children that used chewing gums containing xylitol and casein phosphopeptide-amorphous calcium phosphate. Conclusion: Children that used the chewing gums after ingestion of an acidic soft drink presented an increase in salivary pH, with the best results in the groups that used chewing gums containing casein phosphopeptide-amorphous calcium phosphate and xylitol.

Particle Competition and Cooperation in Networks for Semi-Supervised Learning

Semi-supervised learning is one of the important topics in machine learning, concerning with pattern classification where only a small subset of data is labeled. In this paper, a new network-based (or graph-based) semi-supervised classification model is proposed. It employs a combined random-greedy walk of particles, with competition and cooperation mechanisms, to propagate class labels to the whole network. Due to the competition mechanism, the proposed model has a local label spreading fashion, i.e., each particle only visits a portion of nodes potentially belonging to it, while it is not allowed to visit those nodes definitely occupied by particles of other classes. In this way, a "divide-and-conquer" effect is naturally embedded in the model. As a result, the proposed model can achieve a good classification rate while exhibiting low computational complexity order in comparison to other network-based semi-supervised algorithms. Computer simulations carried out for synthetic and real-world data sets provide a numeric quantification of the performance of the method.

Titanium surface topography after brushing with fluoride and fluoride-free toothpaste simulating 10 years of use

Objectives: To conduct a controlled study contrasting titanium surface topography after procedures that simulated 10 years of brushing using toothpastes with or without fluoride. Methods: Commercially pure titanium (cp Ti) and Ti-6Al-4V disks (6 mm circle divide x 4 mm) were mirror-polished and treated according to 6 groups (n = 6) as a function of immersion (I) or brushing (B) using deionised water (W), fluoride-free toothpaste (T) and fluoride toothpaste (FT). Surface topography was evaluated at baseline (pretreatment) and post-treatment, using atomic force microscope in order to obtain three-dimensional images and mean roughness. Specimens submitted to immersion were submerged in the vehicles without brushing. For brushed specimens, procedures were conducted using a linear brushing machine with a soft-bristled toothbrush. Immersion and brushing were performed for 244 h. IFT and BFT samples were analysed under scanning electron microscope with Energy-Dispersive X-ray Spectroscopy (EDS). Pre and post-treatment values were compared using the paired Student T-test (alpha = .05). Intergroup comparisons were conducted using one-way ANOVA with Tukey post-test (alpha = .05). Results: cp Ti mean roughness (in nanometers) comparing pre and post-treatment were: IW, 2.29 +/- 0.55/2.33 +/- 0.17; IT, 2.24 +/- 0.46/2.02 +/- 0.38; IFT, 2.22 +/- 0.53/1.95 +/- 0.36; BW, 2.22 +/- 0.42/3.76 +/- 0.45; BT, 2.27 +/- 0.55/16.05 +/- 3.25; BFT, 2.27 +/- 0.51/22.39 +/- 5.07. Mean roughness (in nanometers) measured in Ti-6Al-4V disks (pre/post-treatment) were: IW, 1.79 +/- 0.25/2.01 +/- 0.25; IT, 1.61 +/- 0.13/1.74 +/- 0.19; IFT, 1.92 +/- 0.39/2.29 +/- 0.51; BW, 2.00 +/- 0.71/2.05 +/- 0.43; BT, 2.37 +/- 0.86/11.17 +/- 2.29; BFT, 1.83 +/- 0.50/15.73 +/- 1.78. No significant differences were seen after immersions (p > .05). Brushing increased the roughness of cp Ti and of Ti-6Al-4V (p < .01); cp Ti had topographic changes after BW, BT and BFT treatments whilst Ti-6Al-4V was significantly different only after BT and BTF. EDS has not detected fluoride or sodium ions on metal surfaces. Conclusions: Exposure to toothpastes (immersion) does not affect titanium per se; their use during brushing affects titanium topography and roughness. The associated effects of toothpaste abrasives and fluorides seem to increase roughness on titanium brushed surfaces. (C) 2012 Elsevier Ltd. All rights reserved.

Semi-supervised learning guided by the modularity measure in complex networks

Semi-supervised learning techniques have gained increasing attention in the machine learning community, as a result of two main factors: (1) the available data is exponentially increasing; (2) the task of data labeling is cumbersome and expensive, involving human experts in the process. In this paper, we propose a network-based semi-supervised learning method inspired by the modularity greedy algorithm, which was originally applied for unsupervised learning. Changes have been made in the process of modularity maximization in a way to adapt the model to propagate labels throughout the network. Furthermore, a network reduction technique is introduced, as well as an extensive analysis of its impact on the network. Computer simulations are performed for artificial and real-world databases, providing a numerical quantitative basis for the performance of the proposed method.

Surface roughness and color change of a composite: Influence of beverages and brushing

This study evaluated the influence of beverages and brushing on the surface roughness(SR) and color change(Delta E) of a composite resin. For this, 120-disks(10 mmx2 mm) of composite resin(Filtek-Z250) were prepared and polished. Initials SR(Ra-mu m) and color(CIELab-system) were measured with rugosimeter and spectrophotometer; specimens were divided into four groups(red wine, soft drink, sugar cane spirit, or artificial saliva=control) and three subgroups(without brushing; brushed with Colgate or with Close-Up). Specimens were immersed in the beverage 5x/day, for 5', over 30 day, being two subgroups brushed(120 strokes/day). Color was measured at 15th day, 30th day and after repolishment; SR at 30th day. Delta E-values were statistically different after immersion in the beverages(p<0.05). Red wine promoted the highest alteration, followed by soft drink=sugar cane spirit and finally saliva. At 30th day, specimens exhibited Delta E higher than 15th day; after repolishing, Delta E was similar to 15th day. Beverages and brushing negatively influenced the SR. Therefore, Delta E and SR can be influenced by beverages and brushing.

Demographic characterization of supervised dog and cat populations in Sao Paulo city, Brazil

This study aimed to estimate the owned dog and cat populations of Sao Paulo city using a complex sample with random selection in two stages. In each administrative district, six census sectors and 20 households in each sampled sector were visited from September 2006 to September 2009, totalizing 11,272 interviews. The human: dog ratio was 4.34 and the human: cat ratio was 19.33. The dog population was 2,507,401 and the cat population, 562,965. The dog population was composted of 52.7% males, while among the cat population it was 45.1%. The proportion of sterilized females (23.4% among dogs and 46.1% among cats) was higher than males (11.4% among dogs and 31.5% among cats). The mean age of dogs was 4.99 years, and for cats, 3.53 years. The proportion of restricted (without access to the street) dogs, 64.4%, was higher than restricted cats, 42.5%. The average number of animals/household was 1.60 for dogs and 1.69 for cats. The animal ownership is associated with cultural factors, therefore the characterization of canine and feline population is essential to implement an adequate animal population management program and zoonosis control.

Semi-supervised dimensionality reduction based on partial least squares for visual analysis of high dimensional data

Dimensionality reduction is employed for visual data analysis as a way to obtaining reduced spaces for high dimensional data or to mapping data directly into 2D or 3D spaces. Although techniques have evolved to improve data segregation on reduced or visual spaces, they have limited capabilities for adjusting the results according to user's knowledge. In this paper, we propose a novel approach to handling both dimensionality reduction and visualization of high dimensional data, taking into account user's input. It employs Partial Least Squares (PLS), a statistical tool to perform retrieval of latent spaces focusing on the discriminability of the data. The method employs a training set for building a highly precise model that can then be applied to a much larger data set very effectively. The reduced data set can be exhibited using various existing visualization techniques. The training data is important to code user's knowledge into the loop. However, this work also devises a strategy for calculating PLS reduced spaces when no training data is available. The approach produces increasingly precise visual mappings as the user feeds back his or her knowledge and is capable of working with small and unbalanced training sets.

Efeito de diferentes gomas de mascar sobre o pH salivar de crianças

Objective: To evaluate the effect of different chewing gum brands on the salivary pH of children with primary dentition. Method: Forty children were selected and assigned to four groups: control (no chewing gum); sugarless chewing gum; chewing gum with casein phosphopeptide-amorphous calcium phosphate; and chewing gum with xylitol. The first saliva collection was made after supervised tooth brushing for stabilization of the oral pH. Next, all children were instructed to drink slowly 100 mL of a cola-based soft drink (Coca-Cola®) and a new saliva collection was made 10 min later. Then, each group chewed on the chewing gum for 5 min and discarded it after this time. Saliva was collected again at 5, 10 and 15 min intervals after start using the chewing gum. Measurement of salivary pH was made with colorimetric test papers and a digital pH-meter. Data were analyzed statistically by analysis of variance and Tukey’s test at a 5% significance level. Results: The use of chewing gums accelerated the increase of salivary pH to considerably alkaline levels after consumption of an acidic beverage, especially within the first minutes. The highest levels were obtained in the groups of children that used chewing gums containing xylitol and casein phosphopeptide-amorphous calcium phosphate. Conclusion: Children that used the chewing gums after ingestion of an acidic soft drink presented an increase in salivary pH, with the best results in the groups that used chewing gums containing casein phosphopeptide-amorphous calcium phosphate and xylitol.

Efficacy of graded activity versus supervised exercises in patients with chronic non-specific low back pain: protocol of a randomised controlled trial

Abstract Background Low back pain is a relevant public health problem, being an important cause of work absenteeism worldwide, as well as affecting the quality of life of sufferers and their individual functional performances. Supervised active physical routines and of cognitive-behavioral therapies are recommended for the treatment of chronic Low back pain, although evidence to support the effectiveness of different techniques is missing. Accordingly, the aim of this study is to contrast the effectiveness of two types of exercises, graded activity or supervised, in decreasing symptoms of chronic low back pain. Methods/design Sample will consist of 66 patients, blindly allocated into one of two groups: 1) Graded activity which, based on an operant approach, will use time-contingent methods aiming to increase participants’ activity levels; 2) Supervised exercise, where participants will be trained for strengthening, stretching, and motor control targeting different muscle groups. Interventions will last one hour, and will happen twice a week for 6 weeks. Outcomes (pain, disability, quality of life, global perceived effect, return to work, physical activity, physical capacity, and kinesiophobia) will be assessed at baseline, at treatment end, and three and six months after treatment end. Data collection will be conducted by an investigator blinded to treatment allocation. Discussion This project describes the randomisation method that will be used to compare the effectiveness of two different treatments for chronic low back pain: graded activity and supervised exercises. Since optimal approach for patients with chronic back pain have yet not been defined based on evidence, good quality studies on the subject are necessary. Trial registration NCT01719276

A semi-supervised classification technique based on interacting forces

Semi-supervised learning is a classification paradigm in which just a few labeled instances are available for the training process. To overcome this small amount of initial label information, the information provided by the unlabeled instances is also considered. In this paper, we propose a nature-inspired semi-supervised learning technique based on attraction forces. Instances are represented as points in a k-dimensional space, and the movement of data points is modeled as a dynamical system. As the system runs, data items with the same label cooperate with each other, and data items with different labels compete among them to attract unlabeled points by applying a specific force function. In this way, all unlabeled data items can be classified when the system reaches its stable state. Stability analysis for the proposed dynamical system is performed and some heuristics are proposed for parameter setting. Simulation results show that the proposed technique achieves good classification results on artificial data sets and is comparable to well-known semi-supervised techniques using benchmark data sets.

Supervised classification of basaltic aggregate particles based on texture properties

The strength and durability of materials produced from aggregates (e.g., concrete bricks, concrete, and ballast) are critically affected by the weathering of the particles, which is closely related to their mineral composition. It is possible to infer the degree of weathering from visual features derived from the surface of the aggregates. By using sound pattern recognition methods, this study shows that the characterization of the visual texture of particles, performed by using texture-related features of gray scale images, allows the effective differentiation between weathered and nonweathered aggregates. The selection of the most discriminative features is also performed by taking into account a feature ranking method. The evaluation of the methodology in the presence of noise suggests that it can be used in stone quarries for automatic detection of weathered materials.

Bluetongue disease: impact of the 2008 vaccination on fertility in supervised dairy herds

In June 2008 the compulsary nationwide vaccination against BTV-8 (Bluetongue virus serotype 8) was started. After a short time, several owners complained about undesirable effects of the vaccination on fertility and milk quality. Data from 47 dairy farms, regularly supervised by herd health practitioners, were analysed in order to clarify a possible connection between vaccination and fertility. Both vaccinations given each cow for basic immunization were evaluated according to their effects on conception rate and pregnancy. In model calculations the first vaccination had no significant effect on the first service conception rate (FCR), the all service conception rate (ACR) and on the abortion rate. The second vaccination led to a significantly reduced FCR when the cow was inseminated within 20 days of being vaccinated and to a significantly worse ACR when inseminated 10 days before or after vaccination. However, these individually established reductions of the insemination rate had only little influence on overall data.