Title Usage possibilities of the cloud products of satellite datasets Author: Bc. Petr Šácha Department: Department of Meteorology and Environment Protection Supervisor: RNDr. Petr Pišoft Ph.D. Supervisor's e-mail address: Petr.Pisoft@mff.cuni.cz Abstract: Cloudiness plays an important role in the global energy and water cycle. In particular, the presence of clouds dominates the planetary albedo and takes part in many climate feedback processes. In this thesis a short informational overview of remote sensing, a description of EUMETSAT, satellites, which it used, its part CM- SAF and a search retrieval of current research is given at first. Then the study is focused on the cloud satellite products, especially on CFC (cloud fractional cover) and CTY (cloud type) products. Data sets of daily averages of these products are compared with the daily averages created from the surface SYNOP observations of the total cloud cover and cloud type in the area of the Czech Republic. In the case of big variances between the two examined datasets of daily averaged cloud coverage, possible causes like the sampling error, dependence on season, localization of the station and the elevation and type of cloudiness, are searched. Several statistic analyses and validation scores are computed. Finally, possibilities of the examined...

Yes / Immunotherapy treatments can be essential sometimes and a waste of valuable resources in other cases, depending on the diagnosis results. Therefore, researchers in immunotherapy need to be updated on the current status of research by exploring: application domains e.g. warts, datasets e.g. immunotherapy, classifiers or algorithms e.g. kNN and software tools. The research objectives were: 1) to study the immunotherapy-related published literature from a supervised machine learning perspective. In addition, to reproduce immunotherapy classifiers reported in research papers. 2) To find gaps and challenges both in publications and practical work, which may be the basis for further research. Immunotherapy, diabetes, cryotherapy, exasens data and ”one unbalanced dataset” are explored. The results are compared with published literature. To address the found gaps in further research: novel experiments, unbalanced studies, focus on effectiveness and a new classifier algorithm are suggested.

No / Certain crimes are difficult to be committed by individuals but carefully organised by group of associates and affiliates loosely connected to each other with a single or small group of individuals coordinating the overall actions. A common starting point in understanding the structural organisation of criminal groups is to identify the criminals and their associates. Situations arise in many criminal datasets where there is no direct connection among the criminals. In this paper, we investigate ties and community structure in crime data in order to understand the operations of both traditional and cyber criminals, as well as to predict the existence of organised criminal networks. Our contributions are twofold: we propose a bipartite network model for inferring hidden ties between actors who initiated an illegal interaction and objects affected by the interaction, we then validate the method in two case studies on pharmaceutical crime and underground forum data using standard network algorithms for structural and community analysis. The vertex level metrics and community analysis results obtained indicate the significance of our work in understanding the operations and structure of organised criminal networks which were not immediately obvious in the data. Identifying these groups and mapping their relationship to one another is essential in making more effective disruption strategies in the future.

Providing high school students a course of study for becoming competent and thorough lifelong independent readers of complex texts was the goal for this thesis. This is accomplished by integrating literary theory that looks beyond just the typical level of analysis often emphasized in many Florida classrooms. If put into use and successful, this curriculum guide will aid Florida teachers in endowing their students with a new level of ability to analyze literature. Research of prior work done in the field of integrating critical theory into high school classrooms was analyzed and synthesized in order to create a larger course of critical theory study to be completed during the senior year of high school in the state of Florida. The curriculum guide acts as a starting point, providing teachers with all the tools necessary to bring literary theory into the high school classroom while maintaining their individual teaching style. The curriculum guide is broken into four distinct units which follow the most common course of Florida twelfth grade study, the English canon, with each chapter addressing two literary theories. The literary theories utilized are: New Criticism, New Historicism, Feminism, Marxism, Reader Response, Psychoanalysis, Structuralism, and Deconstruction.

Traditional clustering is typically based on a single feature set. In some domains, several feature sets may be available to represent the same objects, but it may not be easy to compute a useful and effective integrated feature set. We hypothesize that clustering individual datasets and then combining them using a suitable ensemble algorithm will yield better quality clusters compared to the individual clustering or clustering based on an integrated feature set. We present two classes of algorithms to address the problem of combining the results of clustering obtained from multiple related datasets where the datasets represent identical or overlapping sets of objects but use different feature sets. One class of algorithms was developed for combining hierarchical clustering generated from multiple datasets and another class of algorithms was developed for combining partitional clustering generated from multiple datasets. The first class of algorithms, called EPaCH, are based on graph-theoretic principles and use the association strengths of objects in the individual cluster hierarchies. The second class of algorithms, called CEMENT, use an EM (Expectation Maximization) approach to progressively refine the individual clusterings until the mutual entropy between them converges toward a maximum. We have applied our methods to the problem of clustering a document collection consisting of journal abstracts from ten different Library of Congress categories. After several natural language preprocessing steps, both syntactic and semantic feature sets were extracted. We present empirical results that include the comparison of our algorithms with several baseline clustering schemes using different cluster validation indices. We also present the results of one-tailed paired emph{T}-tests performed on cluster qualities. Our methods are shown to yield higher quality clusters than the baseline clustering schemes that include the clustering based on individual feature sets and clustering based on concatenated feature sets. When the sets of objects represented in two datasets are overlapping but not identical, our algorithms outperform all baseline methods for all indices.

No description available.

No description available.

Yes / Real-world datasets are increasingly heterogeneous, showing a mixture of numerical, categorical and other feature types. The main challenge for mining heterogeneous datasets is how to deal with heterogeneity present in the dataset records. Although some existing classifiers (such as decision trees) can handle heterogeneous data in specific circumstances, the performance of such models may be still improved, because heterogeneity involves specific adjustments to similarity measurements and calculations. Moreover, heterogeneous data is still treated inconsistently and in ad-hoc manner. In this paper, we study the problem of heterogeneous data classification: our purpose is to use heterogeneity as a positive feature of the data classification effort by using consistently the similarity between data objects. We address the heterogeneity issue by studying the impact of mixing data types in the calculation of data objects’ similarity. To reach our goal, we propose an algorithm to divide the initial data records based on pairwise similarity for classification subtasks with the aim to increase the quality of the data subsets and apply specialized classifier models on them. The performance of the proposed approach is evaluated on 10 publicly available heterogeneous data sets. The results show that the models achieve better performance for heterogeneous datasets when using the proposed similarity process.

Recent advances in deep generative models have enabled the generation of realistic synthetic media or deepfakes, including synthetic images, videos, and text. However, synthetic media can be misused for malicious purposes and damage users' trust in online content. This dissertation aims to address several key challenges in defending against the misuse of synthetic media. Key contributions of this dissertation include the following: (1) Understanding challenges with the real-world applicability of existing synthetic media defenses. We curate synthetic videos and text from the wild, i.e., the Internet community, and assess the effectiveness of state-of-the-art defenses on synthetic content in the wild. In addition, we propose practical low-cost adversarial attacks, and systematically measure the adversarial robustness of existing defenses. Our findings reveal that most defenses show significant degradation in performance under real-world detection scenarios, which leads to the second thread of my work: (2) Building detection schemes with improved generalization performance and robustness for synthetic content. Most existing synthetic image detection schemes are highly content-specific, e.g., designed for only human faces, thus limiting their applicability. I propose an unsupervised content-agnostic detection scheme called NoiseScope, which does not require a priori access to synthetic images and is applicable to a wide variety of generative models, i.e., GANs. NoiseScope is also resilient against a range of countermeasures conducted by a knowledgeable attacker. For the text modality, our study reveals that state-of-the-art defenses that mine sequential patterns in the text using Transformer models are vulnerable to simple evasion schemes. We conduct further exploration towards enhancing the robustness of synthetic text detection by leveraging semantic features. / Doctor of Philosophy / Recent advances in deep generative models have enabled the generation of realistic synthetic media or deepfakes, including synthetic images, videos, and text. However, synthetic media can be misused for malicious purposes and damage users' trust in online content. This dissertation aims to address several key challenges in defending against the misuse of synthetic media. Key contributions of this dissertation include the following: (1) Understanding challenges with the real-world applicability of existing synthetic media defenses. We curate synthetic videos and text from the Internet community, and assess the effectiveness of state-of-the-art defenses on the collected datasets. In addition, we systematically measure the robustness of existing defenses by designing practical low-cost attacks, such as changing the configuration of generative models. Our findings reveal that most defenses show significant degradation in performance under real-world detection scenarios, which leads to the second thread of my work: (2) Building detection schemes with improved generalization performance and robustness for synthetic content. Many existing synthetic image detection schemes make decisions by looking for anomalous patterns in a specific type of high-level content, e.g., human faces, thus limiting their applicability. I propose a blind content-agnostic detection scheme called NoiseScope, which does not require synthetic images for training, and is applicable to a wide variety of generative models. For the text modality, our study reveals that state-of-the-art defenses that mine sequential patterns in the text using Transformer models are not robust against simple attacks. We conduct further exploration towards enhancing the robustness of synthetic text detection by leveraging semantic features.

Eggplant grafting implemented implemented in two field sites in the Philippines, in Nueva Ecija and Pangasinan are used as proofs of concept to illustrate and validate the feasibility of an impact assessment framework for determining the nutritional impact of technology-oriented agricultural activities. Nutritional impacts are assessed by disaggregating the market demand curve into demand curves by income groups using their separate price elasticities of demand. Considering only price effects, the increase in yields following a per unit cost reduction due to eggplant grafting has positive effects on the daily caloric intake per capita in the different income classes with the greatest impact on the lowest income class for both sites. Net increases in calorie intake ranges between 0.09 and 0.6 kilocalories per capita per day. / Master of Science