2.2 Estudios de maestría
URI permanente para esta colección
Examinar
Examinando 2.2 Estudios de maestría por browse.metadata.type "info:eu-repo/semantics/conferenceObject"
Resultados por página
Opciones de clasificación
-
PublicaciónA deep learning approach for sentiment analysis in Spanish Tweets(Springer Verlag, 2018)
;Vizcarra G. ;Mauricio A.Mauricio L.Sentiment Analysis at Document Level is a well-known problem in Natural Language Processing (NLP), being considered as a reference in NLP, over which new architectures and models are tested in order to compare metrics that are also referents in other issues. This problem has been solved in good enough terms for English language, but its metrics are still quite low in other languages. In addition, architectures which are successful in a language do not necessarily works in another. In the case of Spanish, data quantity and quality become a problem during data preparation and architecture design, due to the few labeled data available including not-textual elements (like emoticons or expressions). This work presents an approach to solve the sentiment analysis problem in Spanish tweets and compares it with the state of art. To do so, a preprocessing algorithm is performed based on interpretation of colloquial expressions and emoticons, and trivial words elimination. Processed sentences turn into matrices using the 3 most successful methods of word embeddings (GloVe, FastText and Word2Vec), then the 3 matrices merge into a 3-channels matrix which is used to feed our CNN-based model. The proposed architecture uses parallel convolution layers as k-grams, by this way the value of each word and their contexts are weighted, to predict the sentiment polarity among 4 possible classes. After several tests, the optimal tuple which improves the accuracy were <1, 2>. Finally, our model presents %61.58 and %71.14 of accuracy in InterTASS and General Corpus respectively. -
PublicaciónA mapping approach for real time imitation of human movements by a 22 DOF humanoid(IEEE, 2018)
;Cornejo-Arismendi V.A.Barrios-Aranibar D.The main way of displacement of a humanoid robot is by walking, humanoid robots have a basic architecture of 22 DOF which are the minimum necessary to replicate human movements. A motion capture system stores the information of a human being from static points in a human body, the data used will be cycles of gait of a human being. The proposed technique transforms the data of a capture system and transforms them into angles in an architecture of a humanoid robot of 22 DOF. For this purpose it uses key points of a capture system and makes a mapping from the torso to then proceed with its upper and lower limbs. Tests were performed on an author's own simulator and also on the V-REP simulator using the architecture of the Poopy robot. The results show a visually imperceptibly mathematical error in the simulator, but numerically measurable, that lies in the elimination of an axial axis located at the waist. Tests were performed with the data of a woman, a man and a child, being the woman who has the greatest error for having a more pronounced hip movement in the gait. This proposed research opens the door for future research that requires a mapping of a capture system to be replicated in a humanoid robot of 22 DOF, being its use very versatile and expandable to dynamic solutions of balance and tightness. -
PublicaciónA visual analytics approach for exploration of high-dimensional time series based on Neighbor-Joining Tree(Association for Computing Machinery, 2018)
;Rodríguez R. ;Alfonte R.Cuadros A.M.High-dimensional time series analysis through visual techniques poses many challenges due to the visualization solutions proposed until now for exploratory tasks are not well-oriented to high volume of data. When the data sets grow large, the visual alternatives do not allow for a good association between similar time series. With the aim to increase more alternatives, we introduce a visual analytic approach based on Neighbor-Joining similarity tree. The proposed approach internally consists of five time series dimension reduction techniques widely used, two well-known similarity measures and interaction mechanisms to do exploratory analysis of high-dimensional time series data interactively. -
PublicaciónAbnormal event detection in video using motion and appearance information(Springer Verlag, 2018)
;Menejes Palomino N.Cámara Chávez G.This paper presents an approach for the detection and localization of abnormal events in pedestrian areas. The goal is to design a model to detect abnormal events in video sequences using motion and appearance information. Motion information is represented through the use of the velocity and acceleration of optical flow and the appearance information is represented by texture and optical flow gradient. Unlike literature methods, our proposed approach provides a general solution to detect both global and local abnormal events. Furthermore, in the detection stage, we propose a classification by local regions. Experimental results on UMN and UCSD datasets confirm that the detection accuracy of our method is comparable to state-of-the-art methods. -
PublicaciónAn enhanced triplet CNN based on body parts for person re-identificacion(IEEE Computer Society, 2018)
;Espinoza J.D. ;Chavez G.C.Torres G.H.Person re-identificacion consists of reidentificating person through a set of images that is taken by different camera views. Despite recent advances in this field, this problem still remains a challenge due to partial occlusions, changes in illumination, variation in human body poses. In this paper, we present an enhanced Triplet CNN based on body-parts for person re-identification (AETCNN). We design a new model able to learn local body-part features and integrate them to produce the final feature representation of each input person. In addition, to avoid over-fitting due to the small size of the dataset, we propose an improvement in triplet assignment to speed up the convergence and improve performance. Experiments show that our approach achieves very promising results in (CUHK01) dataset and we advance state of the art, improving most of the results of the state of the art with a simpler architecture, achieving 76.50% in rank 1. -
PublicaciónAnalytical Study of the Thermal Activation of Tb Doped Amorphous SiC:H Thin Films(Materials Research Society, 2016)
;Guerra Torres, Jorge Andrés ;Tucto K. ;Montañez L.M. ;De Zela F. ;Töfflinger J.A. ;Winnaker A.Weingärtner R.The luminescence of Tb-doped a-SiC:H thin films with different Tb concentrations under sub-bandgap photon excitation was investigated. Two independent processes were identified. The annealing induced activation of the Tb3+ and the inhibition of host-mediated non-radiative recombination paths. The integrated emission intensity is described by a rate equation model, considering these two. In this study, the luminescence enhancement with increasing annealing temperature is shown. The optimal Tb concentration and annealing temperature for the highest Tb-related light emission intensity is determined. Finally, a parameter proportional to the number of optically active ions is found through the aforementioned model. -
PublicaciónApplication of the KDD process for the visualization of integrated geo-referenced textual data from the pre-processing phase(Springer Verlag, 2018)
;Gomez F. ;Iquira D.Cuadros A.M.Geo-referenced textual data has been the subject of multiple investigations, by providing opportunities to better understand certain phenomena according to the content that is shared, either on-line such as social networks, blogs, and news; or through repositories such as scientific research articles, geo-referenced virtual books, among others. However, the characteristics of this information are studied, analyzed and processed separately, either through its textual components or its geo-spatial components, which offers a separate understanding of the results. In this paper, we propose an integration of textual and geo-spatial components from the pre-processing phase to the visualization stage, As a part of the Document Mapping process based on the phases of the Knowledge Discovery in Databases (KDD). Achieving two main results (1) minimize the problems that arise in the visual phase, such as data occlusion and (2) provide a more detailed understanding between the textual relationships of the data when plotted in a geo-spatial map. -
PublicaciónCalibration of remote nanothermometers using nanoparticles of NaYF 4 : Er 3+ , Yb 3+ , Nd 3+(Institute of Physics Publishing, 2018)
;Vásquez Mazzotti D. ;Alvarado L. ;Puga R.Loro H.The nanoparticles of NaYF4: Er3+, Yb3+, Nd3+ present fluorescent emissions by upconversion. These emissions are associated with the doping ions of Er, Yb and Nd in the NaYF4 nanoparticles. The nanoparticles of NaYF4 doped with Er3+ and Yb3+ ions produce upconversion radiation when excited by 980 nm wavelength laser. By incorporating the Nd3+ ions, the NaYF4: Er3+, Yb3+, Nd3+ arrangement is formed, which produces the same excitation with an 880 nm wavelength laser, the product of a cross-transfer process between the Nd, the Yb and the Er. For applications in human tissue it is preferable to use a 880 nm wavelength laser. These nanoparticles of NaYF4: Er3+, Yb3+, Nd3+ have been synthesized in our laboratories with Y (79,3%), Yb (18%), Er (2%) and Nd (0,7%). In this work we have studied the fluorescence behavior of these nanoparticles, which correspond to the transitions 2H11/2 → 4I15/2 y4S3/2 → 4I15/2, and that can be used in the calibration of a remote nanothermometer, because they have a sensitivity to very small changes in temperature. -
PublicaciónComparing Virtual Reality and 2-Dimensional Drawings for the Visualization of a Construction Project(American Society of Civil Engineers, 2019)
;Calderon-Hernandez, C ;Paes, D ;Irizarry, JBrioso, XIn recent years, advanced visualization tools for the representation of construction projects has gained increased attention from educators and industry practitioners. As they become more receptive to such technologies, it is important to quantify and compare the impact of these technologies on users’ visual cognition of building representations. This study performs a quantitative comparison of users’ cognitive performance when using 2-dimensional (2D) drawings of a structure versus using a BIM-based immersive virtual reality (VR) model of the same structure. Participants were asked to complete visual tasks using the VR platform and the 2D drawings. Findings suggest that visual information depicted in VR is more effective in terms of one’s perception accuracy and memory. The study concludes that the immersive VR tool is suitable for architecture, engineering, and construction (AEC) students and professionals, even those with low spatial ability, facilitating learning and enhancing conventional 2D representations. -
PublicaciónDDoS attack detection mechanism in the application layer using user features(Institute of Electrical and Electronics Engineers Inc., 2018)
;Bravo S.Mauricio D.DDoS attacks are one of the most damaging computer aggressions of recent times. Attackers send large number of requests to saturate a victim machine and it stops providing its services to legitimate users. In general attacks are directed to the network layer and the application layer, the latter has been increasing due mainly to its easy execution and difficult detection. The present work proposes a low cost detection approach that uses the characteristics of the Web User for the detection of attacks. To do this, the features are extracted in real time using functions designed in PHP and JavaScript. They are evaluated by an order 1 classifier to differentiate a real user from a DDoS attack. A real user is identified by making requests interacting with the computer system, while DDoS attacks are requests sent by robots to overload the system with indiscriminate requests. The tests were executed on a computer system using requests from real users and attacks using the LOIC, OWASP and GoldenEye tools. The results show that the proposed method has a detection efficiency of 100%, and that the characteristics of the web user allow to differentiate between a real user and a robot. -
PublicaciónEvaluation of a Gamified 3D Virtual Reality System to Enhance the Understanding of Movement in Physics(CSEDU 2017 - 9th International Conference on Computer Supported Education, 2017)
;Becerra, DAI ;Quispe, JAH ;Aceituno, RGA ;Vargas, GMP ;Zamora, FGF ;Mango, JLH ;Figueroa, GPA ;Vizcarra, AAPChana, JWTThe creation of new technological tools in education provides different learning opportunities to students. The present research evaluates an application that we have developed for the use of virtual reality to enhance the understanding of movement in physics, using gamification techniques on the application allowed us to improve the motivation of the students to learn, the validation of this research was made using a methodology to evaluate the didactic value of educational software, and this evaluation was carried out on a group of teachers. -
PublicaciónExtracting Visual Encodings from Map Chart Images with Color-Encoded Scalar Values(Institute of Electrical and Electronics Engineers Inc., 2019)
;Mayhua A. ;Gomez-Nieto E. ;Heer J.Poco J.Map charts are used in diverse domains to show geographic data (e.g., climate research, oceanography, business analysis, etc.). These charts can be found in news articles, scientific papers, and on the Web. However, many map charts are available only as bitmap images, hindering machine interpretation of the visualized data for indexing and reuse. We propose a pipeline to recover both the visual encodings and underlying data from bitmap images of geographic maps with color-encoded scalar values. We evaluate our results using map images from scientific documents, achieving high accuracy along each step of our proposal. In addition, we present two applications: data extraction and map reprojection to enable improved visual representations of map charts. -
PublicaciónFP-AK-QIEA-R seed to help the convergence of PECO(Institute of Electrical and Electronics Engineers Inc., 2018)Chire Saire J.E.FP-AK-QIEA-ℝ has been used in Benchmark Functions, Protein Folding Problem and multi-objective problems. PECO was a proposed algorithm with good performance in Protein Folding Problem. The objective of this paper is perform experiments to improve the performance of PECO using FP-AK-QIEA-ℝ to initialize PECO. After plotting the best of ever, a meaningful change on the convergence is observed showing the modification of the convergence and after plotting the number of habitats is noticed a reduction of the number of habitats. In conclusion, adding FP-AK-QIEA-R solution improves the performance of PECO.
-
PublicaciónHigh-resolution generative adversarial neural networks applied to histological images generation(Springer Verlag, 2018)
;Mauricio A. ;López J. ;Huauya R.Diaz J.For many years, synthesizing photo-realistic images has been a highly relevant task due to its multiple applications from aesthetic or artistic [19] to medical purposes [1, 6, 21]. Related to the medical area, this application has had greater impact because most classification or diagnostic algorithms require a significant amount of highly specialized images for their training yet obtaining them is not easy at all. To solve this problem, many works analyze and interpret images of a specific topic in order to obtain a statistical correlation between the variables that define it. By this way, any set of variables close to the map generated in the previous analysis represents a similar image. Deep learning based methods have allowed the automatic extraction of feature maps which has helped in the design of more robust models photo-realistic image synthesis. This work focuses on obtaining the best feature maps for automatic generation of synthetic histological images. To do so, we propose a Generative Adversarial Networks (GANs) [8] to generate the new sample distribution using the feature maps obtained by an autoencoder [14, 20] as latent space instead of a completely random one. To corroborate our results, we present the generated images against the real ones and their respective results using different types of autoencoder to obtain the feature maps. -
PublicaciónInfluence of substrate temperature on the properties of Ag-C films produced by DC Magnetron Sputtering(IOP, 2018)
;Calderon, NZ ;Ampuero, JL ;La Rosa-Toro, APujada, BRSilver-carbon films (Ag-C), rich in silver, deposited by DC magnetron sputtering from a silver target (Ag) into silicon substrate (100) in atmosphere of argon/acetylene plasma, have been studied in function of the substrate temperature (as-deposited, 200°C and 300°C). The Ag-C films were obtained using an acetylene gas flow of 0.3 sccm and dc cathode current of 180mA. Ag-C films were characterized by X-ray diffraction, Raman spectroscopy and scanning electron microscopy. It has been observed the decrease of graphite-like bonds in the Ag-C films with the substrate temperature. The SEM revealed that the substrate temperature promotes changes on the surface morphology. These changes on the chemistry and surface morphology are discussed in terms of atoms diffusion on the surface and the chemical composition of the Ag-C films. -
PublicaciónMethodological framework to integrate social and physical vulnerability in the prevention of seismic risk(WITPress, 2018)
;Izquierdo-Horna L.Kahhat R.The huge variety of natural disasters that have occurred in Peru highlights the necessity, due to the social, economic and environmental impacts, to improve national risk and disaster management plans. Hence the need to evaluate vulnerability from a multidimensional perspective that allows the assessment of the social and physical conditions of the region. The main objective of this research is to help improve risk and disaster management plans by proposing a methodology that integrates the assessment of physical vulnerability (based on the physical characteristics of the infrastructure) with social vulnerability, as a result of the main sociodemographic variables. Using the Peruvian city of Chiclayo (located in the district of Chiclayo) as a case study, this methodology has been implemented for the specific case of earthquakes. In order to determine the overall vulnerability of this district, it was first necessary to determine each component, social and physical, separately. For the physical aspect of vulnerability, the amount of debris generated after a seismic event was used as a proxy. For the social aspect, representative and valid social indicators were used for the study area, such as level of education, age profile, permanent disability. By integrating both results, we obtain a global perspective on the vulnerability, giving equal importance, in the assessment of seismic risk, to the individual and infrastructure. The final result of this methodological tool is identification of which sectors need immediate help after a seismic event. Otherwise, if the seismic risk assessment is based only on information about physical aspects, certain areas of the district, that require urgent actions, may be neglected. Finally, this methodology is applicable for the different scales of analysis (e.g. country, region and district). -
PublicaciónMolecular identification of iron oxidizing bacteria isolated from acid mine drainages in Peru(Scientific.net, 2013)
;Abanto, M ;Pariona, N ;Calderon, J ;Guerra, G ;Ramirez, R ;Delacruz-Calvo, FRamirez, PAcidophilic iron-oxidizing microorganisms are important in both environmental and biotechnological applications. These microorganisms are known to accelerate the dissolution of sulfur minerals such as pyrite (FeS2), leading to the acid mine drainage generation , a serious pollution problem, that makes these microorganisms essential to the commercial processing of minerals and sulfur. In order to answer this question, diversity of native acidophilic bacteria isolated from acid mine drainage of Peru was evaluated. The samples were collected from Yanacocha mining (3000 m.a.s.l.) located in the North of Cajamarca region, Yanamina mining (4440 m.a.s.l.) located in the middle of Huancavelica region; finally, SPCC mining (2000 m.a.s.l.) located in the South of Moquegua region. We isolated 11 strains from which three were identified as Acidithiobacillus ferrooxidans, two as At. ferrivorans, two as At. ferridurans, three as Leptospirillum ferrooxidans and one as Acidiphilium sp. by comparative sequencing of PCR-amplified 16S rRNA genes. Phylogenetic analysis of the 16S rRNA genes revealed that some of the strains isolated are closely related to other already known, but there are some with similarities lower than < 95 percent. Our results provide the first study on the diversity of iron-oxidizing bacteria isolated from acid mine drainage of Peru. -
PublicaciónNon-rigid 3D Shape Classification based on Convolutional Neural Networks(IEEE, 2017)
;Quenaya, JFLDel Alamo, CJLOver the years, the scientific interest towards 3D models analysis has become more popular. Problems such as classification, retrieval and matching are studied with the idea to offer robust solutions. This paper introduces a 3D object classification method for non-rigid shapes, based on the detection of key points, the use of spectral descriptors and deep learning techniques. We adopt an approach of converting the models into a “spectral image”. By extracting interest points and calculating three types of spectral descriptors (HKS, WKS and GISIF), we generate a three-channel input to a convolutional neural network. This CNN is trained to automatically learn features such as topology of 3D models. The results are evaluated and analyzed using the Non-Rigid Classification Benchmark SHREC 2011. Our proposal shows promising results in classification tasks compared to other methods, and also it is robust under several types of transformations. -
PublicaciónSeismic Evaluation of Incremental Seismic Retrofitting Techniques for Typical Peruvian Schools(ASCE, 2017)
;Loa, G ;Munoz, ASanta-Cruz, STypical Peruvian schools were built with a now out-of-date seismic code that does not guarantee the student population safety. This type of buildings presents an undesirable structural failure known as short column, i.e. slender columns partially restrained by infill walls. The objective of this paper is to evaluate a low cost retrofitting techniques proposed by a previous study in accordance to the incremental retrofitting methodology proposed by FEMA 395. Incremental dynamic analysis, IDA is used to analyze original and retrofitted structure in terms of structural response and damage state for different intensity levels. The results of these analyses reveal the effectiveness of the interventions in terms of economic values. -
PublicaciónSimilarity-based visual exploration of very large georeferenced multidimensional datasets(Association for Computing Machinery, 2019)
;Peralta-Aranibar R. ;Comba J.L.D. ;Pahins C.A.L.Gomez-Nieto E.Big data visualization is a main task for data analysis. Due to its complexity in terms of volume and variety, very large datasets are unable to be queried for similarities among entries in traditional Database Management Systems. In this paper, we propose an effective approach for indexing millions of elements with the purpose of performing single and multiple visual similarity queries on multidimensional data associated with geographical locations. Our approach makes use of Z-Curve algorithm to map into 1D space considering similarities between data. Additionally, we present a set of results using real data of different sources and we analyze the insights obtained from the interactive exploration.