The main purpose of this paper is to demonstrate the data reduction technique of self-organizing maps and to compare it with data reduction techniques in SPSS. Especially, factor analysis and multidimensional scaling (MDS) are chosen. Subsequent to data reduction a cluster analysis was conducted. Due to taking the same cluster algorithm on the base of different data reduction approaches we can compare the final outputs of the cluster algorithm in respect to a target criterion. This is the homogeneity within the groups compared to the homogeneity between the groups. The application example is taken from literature (Backhaus et al. 1994).

Concern about feeding an ever increasing population has long been one of humankind’s most pressing problems. This has been addressed throughout history by introducing into farming systems changes allowing them to produce more per unit of land area. However, these changes have also been linked to negative effects on the socio economic and environmental sphere, that have created the need for an integral understanding of this phenomenon. This thesis describes the application of learning machine methods to induct a relationship between the spectral response of farms’ land cover and their intensification levels from a sample of farming of Urdaneta municipality, Aragua state of Venezuela. Data collection like this is a necessary first steep to implement cost-effective methods that can help policymakers to conduct succesful planing tasks, especially in countries such as Venezuela where, in spite of there being areas capable of agricultural production, nearly 50% of the internal food requirements of recent years have been satisfied by importations. In this work, farm intensification levels are investigated through a sample of farms of Urdaneta Municipality, Aragua state of Venezuela. This area is characterised by a wide diversity of farming systems ranging from crop to crop-livestock systems and an increasing population density in regions capable of livestock and arable farming, making it a representative case of the main tropical rural zones. The methodology applied can be divided into two main phases. First an unsupervised classification was performed by applying principal component analysis and agglomerative cluster methods to a set of land use and land management indicators, with the aim to segregate farms into homogeneous groups from the intensification point of view. This procedure resulted in three clusters which were named extensive, semi-intensive and intensive. The land use indicators included the percentage area within each farm devoted to annual crops, orchard and pasture, while the land management indicators were percentage of cultivated land under irrigation, stocking rate, machinery and equipment index and permanent and temporary staff ratio, all of them built from data held on the 1996- 1997 venezuelan agricultural census. The previous clusters reached were compared to the ones obtained by applying the learning machine method known as self-organizing map, which is also an unsupervised classification technique, as a way to confirm the groups’ existence. In the second stage, the learning machine known as kernel adatron algorithm was implemented seeking to identify the intensification level of Urdaneta farms from a landsat image, which consisted of two sequential steps: namely training and validation. In the training step, a predetermined number of instances randomly selected from the data set were analysed looking for a pattern to establish a relationship between the label and the spectral response in an iterative process which was concluded when the machine found a linear function capable of separating the two classes with a maximum margin. The supervised classification finishes with the validation in which the kernel adatron classifies the unseen samples by using a generalisation of the relationships learned while training. Results suggest that farm intensification levels can be effectively derived from multi-spectral data by adopting a machine learning approach like the one described.

Increasing interest for classification of 3D geometrical data has led to discov- ery of PointNet, which is a neural network architecture capable of processing un- ordered point sets. This thesis explores several methods of utilizing conventional point features within PointNet and their impact on classification. Classification performance of the presented methods was experimentally evaluated and com- pared with a baseline PointNet model on four different datasets. The results of the experiments suggest that some of the considered features can improve clas- sification effectiveness of PointNet on difficult datasets with objects that are not aligned into canonical orientation. In particular, the well known spin image rep- resentations can be employed successfully and reliably within PointNet. Further- more, a feature-based alternative to spatial transformer, which is a sub-network of PointNet responsible for aligning misaligned objects into canonical orientation, have been introduced. Additional experiments demonstrate that the alternative might be competitive with spatial transformer on challenging datasets. 1

This thesis is concerned with future trend prediction on capital markets on the basis of neural networks. Usage of convolutional and recurrent neural networks, Elliott wave theory and scalograms for capital market's future trend prediction is discussed. The aim of this thesis is to propose a novel approach to future trend prediction based on Elliott's wave theory. The proposed approach will be based on the principle of classification of chosen patterns from Elliott's theory by the way of convolutional neural network. To this end scalograms of the chosen Elliott patterns will be created through application of continuous wavelet transform on parts of historical time series of price for chosen stocks.

<p>Istraživanja predstavljena u disertaciji imala su za cilj razvoj nove metode bazirane na dubokim konvolucijskim neuoronskim mrežama u cilju detekcije bolesti biljaka na osnovu slike lista. U okviru eksperimentalnog dela rada prikazani su dosadašnji literaturno dostupni pristupi u automatskoj detekciji bolesti biljaka kao i ograničenja ovako dobijenih modela kada se koriste u prirodnim uslovima. U okviru disertacije uvedena je nova baza slika listova, trenutno najveća po broju slika u poređenju sa javno dostupnim bazama, potvrđeni su novi pristupi augmentacije bazirani na GAN arhitekturi nad slikama listova uz novi specijalizovani dvo-koračni pristup kao potencijalni odgovor na nedostatke postojećih rešenja.</p> / <p>The research presented in this thesis was aimed at developing a novel method based on deep convolutional neural networks for automated plant disease detection. Based on current available literature, specialized two-phased deep neural network method introduced in the experimental part of thesis solves the limitations of state-of-the-art plant disease detection methods and provides the possibility for a practical usage of the newly developed model. In addition, a new dataset was introduced, that has more images of leaves than other publicly available datasets, also GAN based augmentation approach on leaves images is experimentally confirmed.</p>

Made available in DSpace on 2014-06-11T19:22:25Z (GMT). No. of bitstreams: 0 Previous issue date: 2006-06-30Bitstream added on 2014-06-13T19:48:44Z : No. of bitstreams: 1 sartori_lr_me_prud.pdf: 1503241 bytes, checksum: 70f9983e4d75d8593ab7f2d397146db7 (MD5) / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / Imagens multiespectrais de alta resolução espacial podem se constituir em uma fonte de dados adequada para o mapeamento de processos associados ao desenvolvimento de culturas agrícolas, como a detecção de plantas daninhas. A aerofotogrametria convencional e imagens de satélite de alta resolução espacial são alternativas para aquisição dessas imagens. Entretanto, devido ao custo elevado da aquisição destas imagens, tem sido desenvolvido, pelo Departamento de Cartografia da Faculdade de Ciências e Tecnologia da UNESP de Presidente Prudente, um Sistema de Sensoriamento Remoto Aerotransportado (SRA), capaz de oferecer resolução espacial sub-métrica. Este trabalho considerou a hipótese de que a partir de imagens adquiridas pelo Sistema é possível discriminar graus de infestação de plantas daninhas em culturas de café. Para investigar esta hipótese, foi realizado o mapeamento de plantas daninhas utilizando dois diferentes métodos: classificação de imagens multiespectrais (classificação por redes neurais artificiais - RNA) e análise geoestatística (krigagem por indicação com dados indiretos). Os mapas temáticos foram submetidos à análise da qualidade temática. A krigagem por indicação apresentou resultados suavizados e confusos, ao contrário da classificação por RNA, a qual se constituiu num método eficiente para o objetivo proposto, confirmando a hipótese inicial da investigação. / Multispectral images with high spatial resolution can be a suitable data source for the mapping of processes associated to the crop development, as detection of weed. The aerial photogrammetry and satellite image of high spatial resolution are alternatives for the aquisition of these images. However, due to the high cost of these images, a Sistema de Sensoriamento Remoto Aerotransportado - SRA, which is capable of to offer sub-metric spatial resolution has been developed by the Department of Cartography, FCT/Unesp (Presidente Prudente). This work taked into account the hypothesis that is possible to discriminate infestation degree of weed in coffee crop from high spatial resolution multispectral images. To investigate this hypothesis, it was accomplished the mapping using two different methods: multispectral images classification (artificial neural networks classification) and geoestatistics analysis (Indicator kriging with soft data). The thematics maps were submitted to the analysis of thematic quality. The indicator kriging showed smoothed and confused results instead of the artificial neural networks classification, whose results were efficient for the purpose, confirming the initial hypothesis of the investigation.

This thesis focuses on comparison of selected data mining methods for solving classification tasks with the method of logistic regression. First part of the thesis briefly introduces data mining as a scientific discipline and classification task is shown in the context of knowledge data discovery. Next part explains the principle of particular methods amongst which, along with logistic regression, artificial neural networks, classification decision trees and Support Vector Machine method were selected. Together with mathematical background of each algorithm, demonstration of how the classification functions for new examples is mentioned. Analytical part of this thesis tests decribed methods on real-world data from the Lending Club company and they are compared based on classification accuracy. Towards the end, an evaluation of logistic regression is made in terms of whether its majority position is due to historical reasons or for its high classification accuracy compared to other methods.

Thesis focuses on analysis of contemporary machine learning methods used for text classification based on emotion and testing several deep neural nework architectures. Outcome of this thesis is a neural network architecture, which is tuned for using with text data and which had the best result of 79,94 percent. Proposed method is language independent and it doesn’t require as precisely classified training datasets as current methods. Training and testing datasets were consisted of short amateur movie reviews in Czech and in English. Thesis contains also overview of theoretical basics for convolutional neural networks and history of neural networks and language processing Scripts were written in Python, neural networks were simulated using Keras library and Theano framework. We used CUDA for better performance.

This paper discusses design and implementation of classifying system for recognition of musical instruments from audio records with use of Musical Information Retrieval techniques. In the first part, paper describes parameters used for instrument classification, calculation of said parameters from records and reduction of feature vector. Next part is devoted to tuning and implementation of various classifiers with focus on neural networks. These classifiers ar further tested on records from IRMAS dataset wchich contain 11 musical instruments playing solo or with other instruments. Results of classifiers tested on different parameters and different numbers of instruments are discussed in the last part.

This diploma thesis deals with the topic of automatic classification of buildings. The main goal of this diploma thesis was to design an algorithm for the identification of building types for the purposes of cartographic generalization. For the purposes of this diploma thesis, a total of six types of development were defined with respect to different generalization of individual types on ZM 50. The first part of the proposed method is represented by an algorithm for segmenting buildings into clusters based on the use of already generalized road network and DBSCAN algorithm. The partial goal of this diploma thesis was to compare classifiers from the field of machine learning and neural networks and at the same time to compare classifiers using descriptive characteristics with classifiers using visual assessment. The resulting classifications were evaluated using data from a manually selected training set and using an algorithm comparing the resulting type of development with the type of cartographic representation used to represent the development on ZM 50. The whole method was implemented in Python using Arcpy, Scikit-learn and Tensorflow libraries. Testing took place on elements from the ZABAGED and Data50 databases. Keywords: Generalization of built-up areas, Classification, Machine learning,...