|
|
Crude Oil Price Forecast based on Text News
Skalický, Jan ; Bojar, Ondřej (advisor) ; Žabokrtský, Zdeněk (referee)
For crude oil price forecast, there is a whole range of algorithms. In this thesis we bring out a new perspective on this issue and introduce our project COPF. Using a maximum entropy classifier, we try to predict the change in crude oil price from text information available on the Internet. We are taking advantage of the knowledge of experts in the field. As a part of the thesis, we tested and improved COPF precision. We have found out that this approach poses a lot of interesting problems. In the current state, the precision of our prediction surpassed the baseline but for further development, it is necessary to obtain more data sources. Our algorithm has never been regarded as a self-standing method but it may nicely complement numerical algorithms.
|
|
|
Dynamics of orchid occurrence in highlands
Černocká, Veronika ; Kindlmann, Pavel (advisor) ; Křenová, Zdeňka (referee)
All species of orchids are sensitive plants and records of their distribution are rare. Their rarity is related to their complex biology and their sensitivity to changes in the environment. It is necessary to pay attention to these threatened species, so that we could preserve them from extinction. For their effective protection we need to know the main causes of extinction of their localities. It is also useful to know the factors, which determine their distribution, so that we could find new localities, where they occur. In this thesis I analyzed some climatic, geographic and ecological factors in the Jeseníky mountains and their surroundings. First, I visited localities of orchids, which were registered in large databases of their presence in the past. The four most numerous species were analyzed using the program MaxEnt. Based on the revisions, existence of 2/3 of the revised localities was confirmed, 8% of the revised localities became extinct because of overgrowth or due to human intervention in the landscape. From the MaxEnt results it is clear, that for all species the most important factor at the 50 × 50 m scale, is the type of biotope and heterogeneity of the countryside. The most suitable biotopes are floodplain and wetland forests, mesophile meadows, natural scrubs, peat bogs and...
|
|
|
Spreading of praying mantis (Mantis religiosa) in Europe
Vitáček, Jakub ; Janšta, Petr (advisor) ; Gvoždík, Václav (referee)
Climate change is one of the most important factor determining species ranges. In Europe there is now evidence for northward areal expansion in many Mediterranean insects including the praying mantis (Mantis religiosa). This species is the only representative of the order Mantodea inhabiting central Europe. The northern edge of the species distribution currently reaches latitude 53ř North. Although, the praying mantis is well known insect there is not enough evidence about its phylogeography. In this work three mitochondrial genes (COI, COII, Cyt b) were selected for phylogenetic study. Results indicate three statistically supported distinct lineages in Europe: Eastern European, Central European and Western European. Presumably these lineages are consistent with isolation during the last glacial and re-colonization from glacial refugia. Reduced haplotype diversity on the northern edge suggests currently established populations at the northern distribution border. To validate mtDNA results it was also considered four microsatellite loci. Due to different type of inheritance mtDNA and nuclear DNA it is possible to compare two independent genetic datasets. Microsatellite analysis confirmed results obtained on mitochondrial data. Three major genetic clusters were found: east, west and central. Spatial...
|
|
|
Habitat requirements and predictive distribution modelling of Huperzia selago
Trachtová, Pavla ; Wild, Jan (advisor) ; Douda, Jan (referee)
Studies of the occurrence of montane and boreomontane species in ravines of the sandstone landscape are scarce and the occurrence of these species are explained by the presence of temperature inversion. The question is, which factors limit the occurrence ofthese species in ravines with temperature inversion. The aim of this diploma thesis is to reveal factors that influence the occurrence of Huperzia selago in inverse ravines of sandstone landscape. This work uses a habitat variables recorded directly for populations of H. selago and variables derived from a digital elevation model. These derived variables are also used for creation of two predictive models of geographic distribution of H. selago in the National Park Bohemian Switzerland. When we summarize the most informative variables of predictive models and habitat conditions significantly different from control sites, we get the typical habitat of H. selago. Such sites will likely be found on the rock at the bottom of the valley. Factors that influence the suitability of habitat are: moisture, vegetation type, slope, and distance to the bottom of the valley.
|
|
|
Crude Oil Price Forecast based on Text News
Skalický, Jan ; Bojar, Ondřej (advisor) ; Žabokrtský, Zdeněk (referee)
For crude oil price forecast, there is a whole range of algorithms. In this thesis we bring out a new perspective on this issue and introduce our project COPF. Using a maximum entropy classifier, we try to predict the change in crude oil price from text information available on the Internet. We are taking advantage of the knowledge of experts in the field. As a part of the thesis, we tested and improved COPF precision. We have found out that this approach poses a lot of interesting problems. In the current state, the precision of our prediction surpassed the baseline but for further development, it is necessary to obtain more data sources. Our algorithm has never been regarded as a self-standing method but it may nicely complement numerical algorithms.
|
|
|
Modelling potential distribution of invasive plants Indigo bush (Amorpha fruticosa) and Goji berry (Lycium barbarum) in the Czech Republic
Müllerová, Soňa ; Moudrý, Vítězslav (advisor) ; Barták, Vojtěch (referee)
Biological invasions are currently considered to be one of the major threats to biodiversity. Invasive species are introduced species naturalized in regions outside their native range with both ecological and economic impact known from many regions all over the world. In the Czech Republic, vascular plants constitute a major group of invasive species. They are characterized by the ability to adapt, spread and occupy new sites. Early identification of potentially dangerous species constitute a possible way to prevent further spread. This bachelor thesis presents the use of species distribution modelling technique as a tool for assessing invasive potential and predicting species invasions. The aim was to create the model of invasive shrub species Indigo bush (Amorpha fruticosa) and Goji berry (Lycium barbarum) potential distribution in the Czech Republic and explore the significance of each predictor variable including variables derived from digital elevation model. Models were created in the resolution of 100 m using Maxent method based on presence-only data and ten predictor variables including three topographic variables: elevation, slope and aspect. The results show that the most significant variables for both species are elevation, land cover which is the most important one for Amorpha fruticosa, and temperature annual range. Slope variable is significant for Lycium barbarum. Aspect is not significant. In terms of AUC discrimination measure, high values were achived for both species. AUC value for Amorpha fruticosa is 0,961 and 0,906 for Lycium barbarum. However, the realism and quality of the model is significantly influenced by the stage of invasion and also by complexity and accuracy of input data.
|
guest ::
