The P M 10 concentrations frequently exceed the legislated air quality standards in this city. Particulate Matter (PM) is a complex mixture of airborne particles that differ in size ( P M 2.5 and P M 10 ), origin and chemical composition. PM is released from natural and anthropogenic sources, such as soils, car exhausts, industry, and power plants therefore increasing PM concentrations in many locations. As the molecules are very large to mix between the water molecules, some pollutants do not dissolve in water . This material is termed as particulate matter and can lead to water pollution PM has been linked to a range of serious respiratory and cardiovascular health problems.
The most important issue ofANN in pollutant forecasting is generalization, which refers to their ability to produce reasonable predictions on data sets other than those used for the estimation of the model parameters , . This issue has an important parameter that should be accounted for, it is data preparation . In this work, the preparation of the data is performed by applying a clustering algorithm. Clustering involves the task of dividing data sets, which assigns the same label to members who belong to the same group, so that each group is more or less homogeneous and distinct from the others. In hard clustering (K-means), data is divided into crisp clusters, where each data set belongs to exactly one cluster. In fuzzy clustering, the data points can belong to more than one cluster, and associated with each of the points are membership grades which indicate the degree to which the data sets belong to the different clusters. For this reason the Fuzzyc-Means clustering algorithm (FCM) was used in order to find relationship among pollutant and meteorological variables. These relationship help us to get additional information that will be used for predicting. Unlike hard classification methods which force data to belong exclusively to one class, FCM allows data to belong to multiple classes with varying degrees of membership.
As to contaminant prediction, it is impossible to turn into a model based on phenomenon description because there are many factors that jointly affect contaminant distribution at any time. It is important to remark that none of the classical approaches succeeds in addressing nonlinear relationships whose complex behaviour either lacks interpretation or is difficult to represent using a straightforward function. Then, in the chemical engineering context the growing number of data available on the Internet or in specialized databases makes the establishment of innovative modelling techniques a significant priority in order to profit from the content of this valuable information. In this context, this paper focuses on defining a method that efficiently allows the modelling of contaminant presence at a key site by meansof predictions based on neural networks.
The period February 21–22, 2016 was characterised by a strong influence of a dust outbreak on the city, which was noted in the data recorded at the measuring sites. The prevailing synoptic meteorological situation in those days confined the Iberian Peninsula to an anticyclonic situation centred on the Azores, as well as in eastern Spain and north- ern Africa. As a result, warm African air masses reached Spain. The situation changed on the 24 with the arrival of cold polar air, affecting the whole of the Iberian Peninsula on February 27. Backward air mass trajectories on Febru- ary 21 (left) and 22 (right) performed with the HYSPLIT model are depicted in Fig. 5. The trajectory associated to 2500-m height presented an African origin and differs from those corresponding to the lower altitudes, with an advection of air masses commencing in the Atlantic and entering the southeast of the Peninsula. On February 22, the trajectory associated to 2500-m height also showed an African origin, and the trajectories linked to lower heights modified their pathway. The 750-m height trajectory is associated with LC, and the 1500-m height trajectory displayed a shorter trajec- tory, originating over the Atlantic Ocean. In addition, there were no air masses of African origin on February 20 and 23.
Pruned models were obtained from fully connected feed-forward neural networks with two hidden units, i.e., there was initially a connection between every unit from a layer and every unit of each consecutive layer. In order to select the best pruned architecture, a validation set was used to compare the net- works. Then, when the best model was ob- tained, the interpretation of the influence of each variable was done in the following way: if an input unit is directly connected to the output unit, then a positive weight means that it is a risk factor as it increases the probability of having depression. Thus, a negative weight means that the variable is a protective factor. Let a hidden unit be connected to the output unit with a positive weight. If an input unit is connected to this hidden unit with a positive value, then the variable represented by this unit is a risk factor. If its weight is negative, then it is a protective factor. On the contrary, if the weight between the hidden unit and the output unit is negative, then a positive value in the connection between the input and the hidden unit means that the variable is a pro- tective factor. Thus, a negative value in the weight that connects the input to the hidden unit means that it is a risk factor. 씰 Table 2 summarizes these influences. This interpre- tation is justified because the hidden units have a hyperbolic tangent as an activation function which delimits its output activation values between –1 and 1.
Air pollution posses significant threats to human health and the environment throughout the developed and devel oping countries. Air pollution is one of the most important environmental problems that is caused by both natural and man-made sources. Major man-made sources of ambient air pollution include industries, transportation, power generation, unplanned urban areas, etc. Therefore, the issue of air quality is receiving more attention as an increasing fraction of the countries popUlation are now living in urban areas and are in demand of a cleaner environment. Air pollutants, once emerged from a variety of sources, are subject to mixing, dis persion, transport and complex series of chemical interaction and physical transformation processes in urban atmospheres
The high values of uranium concentrations in the plant indicate that there was a shift in the material, and radioactive contamination in the northwest section of the plant. Research indicates that once the plant was closed, cleansed tanks or storage hoppers, were used to carry out the separation process of the minerals of interest. This cleaning was done with pressure water hoses, so that all dissolved, contaminated material ran toward plant areas of lower height, resulting in stagnation, and later, resulting in the wind spreading throughout the city . Figure 2 shows the specific activity of uranium in the areas surrounding the plant. The specific activity values reached nearly 2000 Bq/kg northwest of the entrance of the plant. The city is only a few hundred meters from the plant. All dimensions are decreasing in height towards the direction of the Chuviscar River.
A feature of simple majority, and other classic voting systems, is that they re- quire individuals to declare dichotomous preferences: they can only declare if an alternative is preferred to another, or if they are indifferent. All kinds of prefer- ence modalities are identified and voters’ opinions are misrepresented. According to Sen [37, p. 162], “... the method of majority decision takes no account of inten- sities of preference, and it is certainly arguable that what matters is not merely the number who prefer x to y and the number who prefer y to x, but also by how much each prefers one alternative to the other”. This idea had already been considered in the 18th Century by the Spanish mathematician J.I. Morales, who in  states that “opinion is not something that can be quantified but rather something which has to be weighed” (see English translation in McLean and Urken [30, p. 204]), or “... majority opinion ... is something which is independent of any fixed number of votes or, which is the same, it has a varying relationship with this figure” (see English translation in McLean and Urken [30, p. 214]).
When Sobel edge detections is applied over Otsu’s method in a dry crop, the results in line detection is very poor as compared with Sobel applied over RFPCM Level. Sobel+Otsu detect one lateral line, while the best performance is given by Prewitt as shown at the Fig. 3 d), but when other edge detection methods are tested over RFPCM Level and RFPCM Cluster, the best result is given by Sobel as depicted in Fig. 3 e) and f). Unlike Otsu’s method, RFPCM Level and direct cluster processing (RFPCM Cluster), always reach their best performance when are used in junction with Sobel.
Plants belonging to the genus Phyllanthus (Phyllanthaceae) are used in traditional medicine for the treatment of numerous diseases. Several Cuban flora species have proven strong antiviral, antioxidant, and antimutagenic properties. However, plant extracts include many different phytocomponents, and therefore they might exert un- desired toxic side effects. In the present work we evaluate the aqueous extracts toxicity of three Cuban endemic Phyllanthus species: P. williamioides Gr., P. chamaecristoides Urb., and P. microdictyus Urb. Cytotoxicity was measured through Bacterial Survival assay and genotoxicity was detected by meansof the SOS Chromotest, both at concentrations ranging from 0.1 to 2 mg/mL andusing Caulobacter crescentus cells as the in vitro experimental model. LC 50 values were not detected at concentrations tested. P. chamaecristoides and P. microdictyus extracts
Cold plasma treatment is a method that was used in hospitals and other industries to sterilize equipment. Recently, this method is being used in the food industry in the fields of conservation and protection from microorganisms (Lee et al. 2005). This novel non-thermal technology has been viewed for years as an effective method to achieve microbial and endogenous enzymes inactivation. In addition, cold plasma sterilization is an eco-friendly process, which is used for food preservation as an alternative to conventional methods such as, heat application and hydrogen peroxide. This method has certain advantages that are important for the food industry including reduced water usage, cost effective, time efficient and lack of chemical residue (Bourke et al. 2017; Thirumdas et al. 2015; Morris et al. 2009). This treatment affects microorganisms present depending on the different conditions applied in the treatment chamber. The effect of this method over the microorganisms present depends on the type of microorganism, the inoculated medium, the exposure method (direct or indirect) and conditions during the exposure to cold plasma (Pignata et al. 2017). The Dielectric Barrier Discharge (DBD) is one of the techniques used to apply cold plasma treatment, this includes normal atmospheric pressure and high voltage. DBD is mainly used in the food industry to achieve microbial decontamination of foodborne pathogens, specially spores (Min et al. 2016). Due to its non-thermal properties, this plasma treatment has the advantage of achieving microbial deactivation at moderate temperatures. This treatment is beneficial for sterilization of high temperature sensitive material (Butscher et al. 2015). The process uses high voltage using two parallel aluminum plates separated by a dielectric barrier. Atmospheric pressure causes high energy levels to be sustained by dielectric barrier; common barrier materials include glass, quartz, polymers and ceramics (Pankaj et al. 2014).
The effects of air pollution on the environment have traditionally been studied in well differentiated scales: local, regional and global. The interest on the local scale has traditionally focused on the effects on human health. Since the 1950s, authorities of some European countries ordered the building of higher stacks, the segregation of the industrial from the residential areas, and prompted a better use of fuels, with the purpose of improving air quality. Since the 1960s to the 1980s, the emission rates of pollutants were individually managed by each country. Since the 1990s, the air quality management in most of Europe is subject to the standards set by the European Commission within the context of the European Union (EU). The EU Directive 2008/50/EC sets limit values for the concentrationsof several air pollutants in the gaseous and particulate matter phases. In Table 1.1, the anthropogenic sources, features and impact of some key pollutants is summarized. The continuous monitoring of the ambient air concentrationsof some of these key pollutants is nowadays one of the tools used for air quality management. In the EU context, this management is performed by each Member State.
Flight test were conducted using Parrot-AR.Drone platform. Communica- tion routines were deloped to send and receive information from the vehicle. A typical orange traffic cone was selected as the object to avoid. We used a VICON motion tracking system 9 to record accurately the trajectory of vehicle with the maximum precision. This information was used for 3D plotting, and no data was used for the control of the aircraft. As mentioned before in this paper, the only information used by the Yaw-controller is the visual information.
This investigation used a three-layer or FFNN for H simulation, where the ﬁrst layer is the input layer representing input variables, the second layer is the hidden layer, and the third layer is the output layer. This topology has proved its ability in modeling many real-world functional problems (Ata, 2015; Piotrowski et al., 2015; Antonopoulos and Antonopoulos, 2017). The selection of hidden neurons is the tricky part in ANN modeling, as it relates to the complexity of the system being modeled. In this study, the optimum numbers of neurons in the hidden layer was determined by a simple trial and errors process. A range of 2 – 80 neurons were evaluated until a minimum acceptable error was achieved between the predicted and observed output. The next step was to choose the transfer functions for the hidden and output layers. In this study, the logistic sigmoid (Logsig) Eq. (2) was used as the transfer function in the hidden layer and the linear transfer function (Pureline) Eq. (3) was applied in the output layer. Table 4 shows the ANN features for each model and station.
The production of the South American blueberry has increased by over 40% in the last decade. However, during storage and shipping, several problems can lead to rejections. This work proposes a pattern recognition method to automatically distinguish stem and calyx ends and detect damaged berries. First, blueberries were imaged under standard conditions to extract color and geometrical features. Second, five algorithms were tested to select the best features to be used in the subsequent evaluation of classification algorithms and cross-validation. The blueberries classes were control, fungally decayed, shriveled, and mechanically damaged. The original 225 features extracted were reduced to 20 or fewer with sequential forward selection. The best classifiers were Support Vector Machine and Linear Discriminant Analysis. Using these classifiers made it possible to successfully distinguish the blueberries‘ orientation in 96.5 % of the cases. For classifying blueberries into the fungally decayed, shriveled, and mechanically damaged classes, the average performances of the classifiers were above 98 %, 93.3 %, and 90 % respectively. All of the experiments were evaluated using external images with 95 % confidence – 10-fold cross-validation. These results are promising because they will allow for the increase in export quality when implemented in production lines.
The marked trend to increase the combined use of multiple types of data for geophy- sical interpretation has lead to an intensive research on joint inversion strategies. In the present work we explore a generalized methodology for joint inversion based on fuzzyc- means clustering (FCM). Unlike structural or petro-physical joint inversion, this does not require any priori property functional. In our scheme, we build an objective function that includes concurrently the individual objective functions for geophysical data sets andfuzzy clustering. Furthermore, it also involves a β parameter that controls the relative weight of the FCM objective function over the overall inversion. We have applied this formulation to a set of sonic log data and magnetic anomaly data to obtain velocity and magnetiza- tion models with well-defined zones for model parameters. In this case, data and model parameters hold linear relationships. The result shows that the cluster and the property parameters can be estimated jointly to yield meaningful models of the subsurface. In a subsequent experiment we implemented this methodology to frequency domain acoustic waveform data, which hold a nonlinear relationship with the property of interest, in con- junction with magnetic anomaly data. Here we used four different strategies to manage the β clustering parameter. Although one of the strategies yields meaningful subsurface models, for this nonlinear case, a general technique to manage the β value remains an open problem.
For the particular case of absorption system mode- lling usingANN, some works can be summarized: va- rious papers have been devoted to the modelling of the thermodynamic properties of the most commonly used fluids in absorption systems (Sözen et al., 2004b; Şencan, 2007; Şencan et al., 2006; Şencan and Kalogirou, 2005; Sözen and Akçayol, 2004a; Sözen et al., 2003). They also show results regarding the system performance using the ANN results. The modelling of a steam fired double effect absorption chiller in a cooling process, used in the pharmaceutical industry, is presented by Manohar et al. (2006). This study also uses an ANN based on external cooling and chilled water temperatures, with good pre- dicting results. Chow et al. (2002) combine a neural net- work and genetic algorithms for the controlled optimization of a direct-fired absorption system. The system-based controlled approach optimizes the use of fuel and electricity for the economical operation of a commercial absorption unit, concluding that considera- ble savings can be achieved. Yung (2007) reported the optimal chiller sequencing in a semiconductor industry by applying ANN to the power consumption data. The results showed that the electricity consumption could be reduced varying the chillers start-up sequencing. More recent papers include: the work performed by Rosiek and Batlles (2010) on the use ofANN to model a solar-assisted air-conditioning system. This system is based on a commercial single-effect LiBr-H 2 O absorp-
In social networking sites, it is useful to receive recommendations about whom to contact or follow. These recommendations not only allow to es- tablish connections with people one might already know in real life, but also with people or users that have similar interests or are potentially interesting. We propose an approach that tackles contact (followee) recommendation in Twitter by meansoffuzzy logic. This fuzzy approach handles recommenda- tion as a link prediction problem and uses three types of similarity between a pair of users: tweet similarity, followee id similarity, and followee tweet sim- ilarity. These similarities are calculated by extracting user profiles. These profiles are, in turn, obtained by considering Twitter as a heterogeneous in- formation network. To test our approach, we crawled a repository of 6,000 users and 2 million tweets, and we measured accuracy by comparing our re- sults with the actual followee lists of the users. These results, which are also compared against the results given by state-of-the-art methods, show a high accuracy. Other advantages of the fuzzy system include a self-explanatory capability and the ability to produce a non-binary friendship value.
meansof the comparison with prototypical situations, are the following: Classification Functions: Include a registry within one of several predefined classes. Clustering Functions: Include a registry within one of several classes (clusters), but unlike the classification, the classes are determined by the own data, by meansof natural groups based on measures of affinity, similarity or probability. Summary Functions: They generate a compact description of a subgroup of data. A simple example like average can be used and the standard deviations for all the fields. Functions of Analysis of sequences: represent sequential patterns, temporary series. The objective is to generate the sequence of states of the process that tries to represent.