In the present work, we investigate the capacity of machine learning to reconstruct simulated large scale surface temperature anomalies given a sparse observation field. Several methods are combined: self-organizing maps and recurrent neural networks of the temporal trajectory. To evaluate our global scale reconstruction, we base our validation on global climate indices time series and EOF analysis. In our experiments, the obtained reconstructions of the global surface temperature anomalies provide a good correlation (over 90%) with the target values when considering scarce available observations sampling about 0.5% of the globe. We reconstruct the surface temperature anomaly fields from 0.05% of total number of data points. We obtain an RMSE of 0.39°C. We further validate the quality of the results calculating a correlation of 0.92, 0.97 and 0.98 between the reconstructed and target indices of AMO, ENSO and IPO. / Klimatsystemet består av olika komponenter inklusive atmosfären, havet och jorden. Som ett öppet system utbyter det hela tiden energi med resten av universum. Det är också ett dynamiskt system vars utveckling kan förutsägas av kända fysiska lagar. Interaktionen mellan dess olika komponenter leder till en så kallad naturlig variation. Denna variabilitet återspeglas i form av svängningslägen, inklusive AMO, ENSO och IPO. För att studera dessa variationer har vi klimatmodeller som representerar de olika krafterna och deras effekt på klimatförändringar på lång sikt. I detta sammanhang är variationerna i det förflutna klimatet särskilt intressanta och tillåter oss en bättre förståelse av klimatförändringar och bättre förutsäga den framtida utvecklingen. Men för att studera det förflutna klimatet eller paleoklimat är den enda tillgängliga informationen endast fullständig under de senaste 150 åren. Innan dess är de enda tillgängliga indikatorerna naturliga, kallad klimatproxy, som trädringar eller iskärnor. Vi kan härleda tidsserier med klimatdata, till exempel temperatur. Denna information är emellertid knappast tillfälligt såväl som över hela världen. Återskapa det globala klimatet från sådana data hanteras fortfarande dåligt. Länken mellan lokal information och global klimat studeras här med hjälp av statistiska metoder, inklusive neurala nätverk. Det långsiktiga målet med denna studie är att bygga en metod för att rekonstruera paleoklimatet från data om klimatproxy, vi fokuserar inledningsvis på rekonstruktionen av ett så kallat perfekt klimat, det vill säga en modell som endast tar hänsyn till naturlig variation, från rumsligt sällsynta tidsserier. De studerade uppgifterna är de från globala yttemperaturutgångar från den havsatmosfärkopplade IPSL-modellen. Uppgifterna förbehandlas för att ta bort säsongens genomsnittliga cykel och omvandlas till temperaturavvikelser. Dessutom väljs rutnätpunkter som representerar information om proxyer pseudo-slumpmässigt, med respekt för den verkliga dispositionen av dessa, övervägande i norr på kontinenterna. Uppgifterna delas upp i träningsdata (150 år), validering (30 år) och testdata (120 år). De metoder som används kombinerar (1) självorganiserande kartor och hierarkisk stigande klassificering, användbara för att producera en reducerad storlek av inmatningsdata, här baserat på tidskorrelationen mellan temperaturutvecklingen under 150 år, (2) ItCompSOM använder korrelationen mellan klasser erhållna genom självorganiserande kartor för att rekonstruera obevakad data, (3) återkommande nervnätverk för att förklara den temporära komponenten i data och förbättra den tidigare rekonstruktionen. Slutligen är definitionen av nya mätvärden nödvändig för att validera de föreslagna modellerna. Utvärderingen av produkterna görs således genom temporär rekonstruktion av AMO, ENSO, IPO klimatlägen samt genom projicering av huvudkomponenterna i analysen av huvudkomponenterna i inputdata. Således konstrueras en reducerad modell av globala temperaturdata baserad på 150 års fullständiga data först, vilket reducerar den rumsliga informationen från 9216 rutnätpunkter till 191 regioner associerade med 1 medelvärde vardera. För att ansluta denna modell till tidssekvenser av sällsynta temperaturer i världen antas det att varje klass som innefattar minst en observerad proxy-data är känd. Rekonstruktionen av globala yttemperaturutvecklingar med ItCompSOM ger en korrelation till indexen på mer än 90% för endast 0,5% av de initiala observationerna. Detta resultat förbättras kraftigt tack vare återkommande nervnätverk, vilket leder till en korrelation av 0,92, 0,97 respektive 0,98 för AMO, ENSO och IPO med endast 0,05% av observationerna. Dessa poäng förklaras med den använda metoden, regionaliseringen hjälper till att koncentrera informationen. Medan 0,5% av rutpunkterna är lika med 43 poäng, om de är korrekt fördelade, representerar de 22% av informationen om regionerna (43 av 191). Dessa mycket uppmuntrande resultat återstår att tillämpas på verkliga klimatproblem, det vill säga med hänsyn till å ena sidan den externa och antropologiska kraften, osäkerheterna relaterade till de verkliga uppgifterna om ombud å andrasidan.

The ever-increasing density of computer storage devices has allowed the average user to store enormous quantities of multimedia content, and a large amount of this content is usually music. Current search techniques for musical content rely on meta-data tags which describe artist, album, year, genre, etc. Query-by-content systems, however, allow users to search based upon the actual acoustical content of the songs. Recent systems have mainly depended upon textual representations of the queries and targets in order to apply common string-matching algorithms and are often confined to a single query style (e.g., humming). These methods also lose much of the information content of the song which limits the ways in which a user may search. We present a query-by-content system which supports querying in several styles using a Self-Organizing Map as its basis. The results from testing our system show that it performs better than random orderings and is, therefore, a viable option for musical query-by-content.

The method of combining multiple geophysical, geological, or geochemical datasets can reveal patterns of otherwise hidden features in the Earth’s crust. This may aid in geological mapping, locating economic mineral deposits and for general anomaly/feature detection. In this study a multidimensional geophysical approach implementing five geophysical datasets is applied using Self-Organizing Maps (SOM), where the main objective is to locate and understand a previously unknown hypothesized fault in Norrbotten, Sweden. The fault is estimated to extend from the Finnish border in the north, across northern Sweden in the N-S direction at a hypothesized length of > 250 km. Self-Organizing Maps is an unsupervised neural network - originally developed by Finnish physicist Teuvo Kohonen - capable of combining any number of datasets and thereby visualize them on a simple two-dimensional map. The datasets used in the analysis were three magnetic derivatives for the x, y and z components, as well as gamma-ray intensity measurements of the 238U, 40K and 232Th radioisotopes. All these variables have been shown to be effective tools for bedrock mapping and geological feature detection and were hence chosen based on these properties. The results revealed the efficiency of the SOM analysis to represent multivariate data on a 2D plane and proved to be a generally good visualization tool for multiple geophysical datasets. There seems to be a relatively sharp difference in geophysical properties between the eastern and western blocks divided by the hypothesized fault, which may indicate the presence of this crustal scale structure. Despite the evidence found in this study, more investigations are needed to verify the existence and nature of the fault, and the results shown here may motivate further projects by providing indications and suggestive evidence for its presence.

The meridional temperature gradient between mid and high latitudes decreases by Arctic amplification. Following this decrease, the circulation in the mid latitudes may change and, therefore, the meridional flux of heat and moisture increases. This might increase the Arctic temperatures even further. A proxy for the vertically integrated atmospheric horizontal energy flux was analyzed using the self-organizing-map (SOM) method. Climate Model Intercomparison Project Phase 5 (CMIP5) model data of the historical and Representative Concentration Pathway 8.5 (RCP8.5) experiments were analyzed to extract horizontal flux patterns. These patterns were analyzed for changes between and within the respective experiments. It was found that the general horizontal flux patterns are reproduced by all models and in all experiments in comparison with reanalyses. By comparing the reanalysis time frame with the respective historical experiments, we found that the general occurrence frequencies of the patterns differ substantially. The results show that the general structure of the flux patterns is not changed when comparing the historical and RCP8.5 results. However, the amplitudes of the fluxes are decreasing. It is suggested that the amplitudes are smaller in the RCP8.5 results compared to the historical results, following a greater meandering of the jet stream, which yields smaller flux amplitudes of the cluster mean.

This thesis aims to cluster financial account time series by extracting global features from the time series and by using two different dimensionality reduction methods, Kohonen Self Organizing Maps and principal component analysis, to cluster the set of the time series by using K-means. The results are then used to further cluster a set of financial services provided by a financial institution, to determine if it is possible to find a set of services which coincide with the time series clusters. The results find several sets of services that are prevalent in the different time series clusters. The resulting method can be used to understand the dynamics between deposits variability and the customers usage of different services and to analyse whether a service is more used in different clusters. / Målet med denna uppsats är att klustra tidsserier över finansiella konton genom att extrahera tidsseriernas karakteristik. För detta används två metoder för att reducera tidsseriernas dimensionalitet, Kohonen Self Organizing Maps och principal komponent analys. Resultatet används sedan för att klustra finansiella tjänster som en kund använder, med syfte att analysera om det existerar ett urval av tjänster som är mer eller mindre förekommande bland olika tidsseriekluster. Resultatet kan användas för att analysera dynamiken mellan kontobehållning och kundens finansiella tjänster, samt om en tjänst är mer förekommande i ett tidsseriekluster.

In this work, different Rotylenchulus reniformis nematode population numbers affecting cotton plants were spectrally classified using Self-Organized Maps. The hyperspectral reflectance of cotton plants affected by different nematode population numbers were analyzed in order to extract information from the signal that would lead to a fieldworthy methodology for predicting nematode population numbers extant in a plant's rhizosphere. Hyperspectral reflectances from both control and field nematode infestations were used in this work. Various feature extraction and dimensionality reduction methods (e.g., PCA, DWT, and SOM-based methods) were used to extract a reduced set of features. These extracted features were then classified using a supervised SOM classification method. Additionally, this work explores the possibility of combining the standard feature extraction methods with self-organized maps to extract a reduced set of features in order to increase classification accuracies.

No description available.

No description available.

No description available.

Several warm winter events have occurred across the Southeast in the past decade, including 2 major events in 2017 and 2018 in Georgia and South Carolina. Plants will begin their spring growth sooner than climatology would suggest and then be damaged by early spring frosts in what is commonly known as a "false spring" event. Some species of plants, like peaches and blueberries, which produce buds early in the season, are just an example of some of the agricultural products more at risk than others. As an important measure of dormancy time in plants, chill hours present a measurement capable of tracking phenological shifts in plants. While a lack of required chill hours can delay spring emergence, intense warm periods can override the chilling hour requirement and induce spring emergence. This project involves training self-organizing maps (SOMs) to identify periods of anomalous winter warming based on a reduced number of chill hours within a 5-day temporal period compared to the period's climatological average. A second SOM is nested in the node that produced the most anomalous events to identify the range of warming that occurs in the most anomalous events, the synoptic setups of these events, and when these occurred. Hourly 2-meter temperature from ERA5 is used to conduct this analysis over a domain centered primarily over South Carolina and Georgia with a temporal period of 1980-2020. Climatological examination of chill hour accumulations in the past 4 decades show an overall decrease in chill hour accumulation across the past decade (2011-2020) Results indicated that periods of higher-than-average temperatures are increasing with time while periods of average or lower than average temperatures are decreasing with time. Both results were statistically significant by Mann-Kendall test. The results of the nested SOMs suggest that an increase in patterns of southerly flow (a common pattern for warmer temperatures) is occurring through time. A third SOM investigating early spring hard freezes was inconclusive but illustrated that some years had more early spring frosts than others independent of how many warmer than average periods occurred in the main winter. The use of SOMs for investigating climatological and synoptic changes in winter and early spring proved successful and effective. Future modifications to these SOMs could be used to identify more trends that exist within these seasons. / Master of Science / Several warm winter events have occurred across the Southeast in the past decade, including 2 major events in 2017 and 2018 in Georgia and South Carolina. Plants will begin their spring growth sooner than climatology would suggest and then be damaged by early spring frosts in what is commonly known as a "false spring" event. Some species of plants, like peaches and blueberries, which produce buds early in the season, are just an example of some of the agricultural products more at risk than others. As an important measure of dormancy time in plants, chill hours present a measurement capable of tracking shifts from normal winter to spring transition in plants. While a lack of required chill hours can delay leaf emergence and spring blooms, intense warm periods can override the chilling hour requirement and induce this spring emergence. This project involves training self-organizing maps (SOMs), a machine learning model, to identify periods of anomalous winter warming based on a reduced number of chill hours within a 5-day temporal period compared to the period's climatological average. A second SOM is nested in the node that produced the most anomalously warm events to identify the range of warming that occurs in the most anomalous events, the large-scale meteorological setups of these events, and when these occurred. Hourly 2-meter temperature from ERA5, a climatological dataset, is used to conduct this analysis over a domain centered primarily over South Carolina and Georgia with a temporal period of 1980-2020. Climatological examination of chill hour accumulations in the past 4 decades show an overall decrease in chill hour accumulation across the past decade (2011-2020) Results indicated that periods of higher-than-average temperatures are increasing with time while periods of average or lower than average temperatures are decreasing with time. Both of these trend findings were statistically significant by Mann-Kendall test. The results of the nested SOMs suggest that an increase in patterns of southerly flow (a common pattern for warmer temperatures) is occurring through time. A third SOM investigating early spring hard freezes (temperatures low enough to cause damage to plant cellular structures) was inconclusive but illustrated that some years had more early spring frosts than others independent of how many warmer than average periods occurred in the main winter. The use of SOMs for investigating climatological and synoptic changes in winter and early spring proved successful and effective. Future modifications to these SOMs could be used to identify more trends that exist within these seasons.