Please read the abstract in the dissertation / Dissertation (MSc (Agric))--University of Pretoria, 2001. / Plant Production and Soil Science / unrestricted

Os resíduos sólidos urbanos (RSU) são particularmente difíceis de gerenciar por se tratar de um resíduo muito heterogêneo em sua composição (resíduos domésticos, de construção civil, de poda de árvore, entre outros). Com isso torna-se necessário o conhecimento das atuais tecnologias de tratamento de resíduos, para que busquem soluções ambientalmente adequadas, economicamente viáveis, socialmente justas e politicamente aceitáveis. Com base nessa premissa e com a necessidade de se estudar novas fontes renováveis de energia na busca por combustíveis alternativos aos combustíveis fósseis, a utilização de resíduos de poda de árvore como fonte energética apresenta aspectos benéficos ao país, desde que viável econômica e tecnicamente. Assim este trabalho analisa a atual situação da destinação dos resíduos de poda de árvore, e apresenta algumas alternativas para seu aproveitamento energético e não energético. Sendo o objetivo principal desta dissertação avaliar o potencial de utilização dos resíduos de poda de árvore para o seu aproveitamento energético por meio dos processos de adensamento de biomassa (peletização e briquetagem), utilizando como base o caso do campus da Universidade de São Paulo na capital. Nos ensaios de peletização foi possível obter pellets com densidade energética elevada, entre 6 GJ/m3 e 8 GJ/m3, o que corresponde a uma densidade energética de 6 a 9 vezes maior que a da poda de árvore in natura. Os briquetes apresentaram uma densidade energética entre 5 GJ/m3 a 6 GJ/m3, ou seja, uma densidade energética de 5 a 6 vezes maior que da poda de árvore in natura. Entretanto os valores obtidos de durabilidade mecânica dos briquetes foram considerados insatisfatórios nas condições estudadas, ao contrário dos pellets, que apresentaram elevados valores. Neste estudo o processo de peletização se mostrou mais viável tecnicamente do que o de briquetagem, visto que os pellets de resíduo de poda de árvore garantiram uma densidade energética elevada e boa resistência mecânica. / Municipal solid wastes (MSW) are particularly difficult to manage because it is a very heterogeneous waste in its composition (household waste, construction, tree pruning, etc). It becomes more than necessary knowledge of current waste treatment technologies, to seek environmentally adequate solutions, economically viable, socially fair and politically acceptable. Based on these assumptions and the need to study new renewable energy sources in the search for alternative fuels to replace fossil fuels, the use of tree pruning waste as an energy source have beneficial aspects to the country, since economic viable and technically. So this paper analyzes the current situation of final disposal of tree pruning waste, and presents some alternatives for its energy and non-energy use. Since the main objective of this work is to evaluate the potential use of tree pruning waste for their energy application through biomass densification processes (pelletizing and briquetting), based on the case of the campus of the University of São Paulo in capital. In the pelleting tests it was possible to obtain pellets with high energy density between 6 GJ/m3 and 8 GJ/m3, which corresponds to an energy density of 6 to 9 times greater than that of tree pruning in nature. The briquettes had an energy density of between 5 GJ/m3 6 GJ/m3, in other words a power density of 5 to 6 times that of tree pruning in nature. However the values obtained from mechanical durability of briquettes were found to be unsatisfactory under the conditions studied, unlike pellets, they showed high values. In this study, the pelleting process is technically more feasible than briquetting. Since the tree pruning waste pellets ensured a high energy density and good mechanical strength.

Os resíduos sólidos urbanos (RSU) são particularmente difíceis de gerenciar por se tratar de um resíduo muito heterogêneo em sua composição (resíduos domésticos, de construção civil, de poda de árvore, entre outros). Com isso torna-se necessário o conhecimento das atuais tecnologias de tratamento de resíduos, para que busquem soluções ambientalmente adequadas, economicamente viáveis, socialmente justas e politicamente aceitáveis. Com base nessa premissa e com a necessidade de se estudar novas fontes renováveis de energia na busca por combustíveis alternativos aos combustíveis fósseis, a utilização de resíduos de poda de árvore como fonte energética apresenta aspectos benéficos ao país, desde que viável econômica e tecnicamente. Assim este trabalho analisa a atual situação da destinação dos resíduos de poda de árvore, e apresenta algumas alternativas para seu aproveitamento energético e não energético. Sendo o objetivo principal desta dissertação avaliar o potencial de utilização dos resíduos de poda de árvore para o seu aproveitamento energético por meio dos processos de adensamento de biomassa (peletização e briquetagem), utilizando como base o caso do campus da Universidade de São Paulo na capital. Nos ensaios de peletização foi possível obter pellets com densidade energética elevada, entre 6 GJ/m3 e 8 GJ/m3, o que corresponde a uma densidade energética de 6 a 9 vezes maior que a da poda de árvore in natura. Os briquetes apresentaram uma densidade energética entre 5 GJ/m3 a 6 GJ/m3, ou seja, uma densidade energética de 5 a 6 vezes maior que da poda de árvore in natura. Entretanto os valores obtidos de durabilidade mecânica dos briquetes foram considerados insatisfatórios nas condições estudadas, ao contrário dos pellets, que apresentaram elevados valores. Neste estudo o processo de peletização se mostrou mais viável tecnicamente do que o de briquetagem, visto que os pellets de resíduo de poda de árvore garantiram uma densidade energética elevada e boa resistência mecânica. / Municipal solid wastes (MSW) are particularly difficult to manage because it is a very heterogeneous waste in its composition (household waste, construction, tree pruning, etc). It becomes more than necessary knowledge of current waste treatment technologies, to seek environmentally adequate solutions, economically viable, socially fair and politically acceptable. Based on these assumptions and the need to study new renewable energy sources in the search for alternative fuels to replace fossil fuels, the use of tree pruning waste as an energy source have beneficial aspects to the country, since economic viable and technically. So this paper analyzes the current situation of final disposal of tree pruning waste, and presents some alternatives for its energy and non-energy use. Since the main objective of this work is to evaluate the potential use of tree pruning waste for their energy application through biomass densification processes (pelletizing and briquetting), based on the case of the campus of the University of São Paulo in capital. In the pelleting tests it was possible to obtain pellets with high energy density between 6 GJ/m3 and 8 GJ/m3, which corresponds to an energy density of 6 to 9 times greater than that of tree pruning in nature. The briquettes had an energy density of between 5 GJ/m3 6 GJ/m3, in other words a power density of 5 to 6 times that of tree pruning in nature. However the values obtained from mechanical durability of briquettes were found to be unsatisfactory under the conditions studied, unlike pellets, they showed high values. In this study, the pelleting process is technically more feasible than briquetting. Since the tree pruning waste pellets ensured a high energy density and good mechanical strength.

隨著知識經濟時代的來臨，掌握知識可幫助組織提昇其競爭力，因此對於知識的產生、儲存、應用和整合，已成為熱烈討論的議題，本研究針對知識整合議題進行探討；而在知識呈現方式中，決策樹（Decision Tree）形式知識為樹狀結構，可以用圖形化的方式來呈現，它的結構簡單且易於瞭解，本研究針對決策樹形式知識來探討其知識整合的課題。 本研究首先提出一個合併選擇決策樹方法MODT（Merging Optional Decision Tree），主要是在原始決策樹結構中增加一個選擇連結（Option Link），來結合具有相同祖先（Ancestor）的兩個子樹；而結合方式是以兩兩合併的方式，由上而下比對兩棵決策樹的節點（Node），利用接枝（Grafting）技術來結合兩棵樹的知識。再者利用強態法則（Strong Pattern Rule）概念來提昇合併樹的預測能力。 其次，由於MODT方法在合併兩棵不同根節點的決策樹時，會形成環狀連結的情形而破壞了原有的樹形結構，以及新增的選擇連結會增加儲存空間且不易維護，因此本研究提出決策樹合併修剪方法DTBMPA（Decision-Tree-Based Merging-Pruning Approach）方法，來改善MODT方法的問題，並且增加修剪程序來簡化合併樹。此方法包括三個主要程序：決策樹合併、合併樹修剪和決策樹驗證。其做法是先將兩棵原始樹經由合併程序結合成一棵合併樹，再透過修剪程序產生修剪樹，最後由驗證程序來評估修剪樹的準確度。本研究提出的DTBMPA方法藉由合併程序來擴大樹的知識，再利用修剪程序來取得更精簡的合併樹。 本研究利用實際信用卡客戶的信用資料來進行驗證。在MODT方法的實驗上，合併樹的準確度同時大於或等於兩棵原始樹的比例為79.5%；並且針對兩者的準確度進行統計檢定，我們發現合併樹的準確度是有顯著大於原始樹。而在DTBMPA方法的實驗中，合併樹的準確度優於原始一棵樹的比率有90%，而修剪樹的準確度大於或等於合併樹的比率有80%。在統計檢定中，合併樹和修剪樹的準確度優於一棵樹的準確度達顯著差異。且修剪樹的節點數較合併樹的節點數平均減少約15%。綜合上述，本研究所提之MODT方法和DTBMPA方法皆能使得合併樹的準確度優於一棵樹的準確度，而其中DTBMPA方法可以取得更精簡的合併樹。 此外，就決策樹形式知識整合的應用而言，本研究提出一個決策樹形式知識發掘預測系統架構，其主要的目在於提供一個Web-Based的知識發掘預測系統，以輔助企業進行知識學習、知識儲存、知識整合、知識流通和知識應用等知識管理的功能。期能藉由使用這套系統來發掘企業內部隱含的重要知識，並運用此發掘的知識進行分類和預測工作。它包含三個主要子系統，即知識學習子系統、合併決策樹子系統和線上預測子系統，其中合併決策樹子系統就是應用本研究所提出之決策樹形式知識整合方法來進行知識整合處理。 有關後續研究方面，可針對下列議題進行研究： 一、就決策樹形式知識整合架構中，探討決策樹形式知識清理單元，即前置處理部份的功能設計，期能讓合併樹結合有一定質量的決策樹形式知識。 二、就綜合多個預測值部份，可加入模糊邏輯理論，處理判定結果值之灰色地帶，以提昇合併樹的預測準確度。 三、就決策樹本身而言，可進一步探討結合選取多個屬性來進行往下分群的決策樹。針對分類性屬性的分支數目不同或可能值不同時的合併處理方法；以及數值性屬性選取不同的分割點時的合併處理方法。 四、探討分類性屬性的分支數目不同或可能值不同時之合併處理方法，以及數值性屬性選取不同的分割點時之合併處理方法。 五、對於合併樹的修剪方法，可考量利用額外修剪例子集來進行修剪的處理方法，並比較不同修剪法之修剪效果及準確度評估。 六、探討多次合併修剪後的決策樹之重整課題，期能藉由調整樹形結構來提昇其使用時的運作效率，且期能讓合併樹順應環境變化而進行其知識調整，並進一步觀察合併樹的樹形結構之變化情形。 七、就實際應用而言，可與廠商合作來建置決策樹形式知識發掘預測系統，配合該廠商的產業特性及業務需求來設計此系統，並導入此系統於企業內部的營運，期能藉此累積該企業的知識且輔助管理者決策的制定。 / In the knowledge economy era, mastering knowledge can improve organization competitive abilities. Therefore, knowledge creation, retention, application, and integration are becoming the hottest themes for discussion nowadays. Our research focuses on the discussion of knowledge integration and related subjects. Decision trees are one of the most common methods of knowledge representation. They show knowledge structure in a tree-shaped graph. Decision trees are simple and easily understood; thus we focus on decision-tree-based knowledge in connection with the theme of knowledge integration. First, this research proposes a method called MODT (Merging Optional Decision Tree), which merges two knowledge trees at once and adds an optional link to merge nodes which have the same ancestor. In MODT, we compare the corresponding nodes of two trees using the top-down traversal method. When their nodes are the same, we recount the number of samples and recalculate the degree of purity. When their nodes are not the same, we add the node of the second tree and its descendants to the first tree by the grafting technique. This yields a completely merged decision tree. The Strong Pattern Rule is used to strengthen the forecast accuracy during the merged decision trees. Secondly, when we use the MODT method to merge two trees which have different roots, the merged tree has cyclic link in the root. It makes the merged tree not a tree structure, so we propose another approach called DTBMPA (Decision-Tree-Based Merging-Pruning Approach) to solve this problem. There are three steps in this approach. In the merging step, the first step, two primitive decision trees are merged as a merged tree to enlarge the knowledge of primitive trees. In the pruning step, the second step, the merged tree from the first step is pruned as a pruned tree to cut off the bias branches of the merged tree. In the validating step, the last step, the performance of the pruned tree from the second step is validated. We took real credit card user data as our sample data. In the MODT experiments, the merged trees showed a 79.5% chance of being equal or more accurate than the primitive trees. This research result supports our proposition that the merged decision tree method could achieve a better outcome with regard to knowledge integration and accumulation. In the DTBMPA simulation experiments, the percentage accuracy for the merged tree will have 90% of chance that is greater than or equal to the accuracy for those primitive trees, and the percentage accuracy for the pruned tree will have 80% of chance that is greater than or equal to the accuracy for merged tree. And we also find that the average number of nodes of the pruned tree will have 15% less than that of the merged tree. Eventually, our MODT and DTBMPA methods can improve the accuracy of the merged tree, and the DTBMPA method can produced smaller merged tree. Finally, in respect to the application of the Decision-Tree-Based Knowledge Integration, this research proposes an on-line Decision-Tree-Based knowledge discovery and predictive system architecture. It can aid businesses to discover their knowledge, store their knowledge, integrate their knowledge, and apply their knowledge to make decisions. It contains three components, including knowledge learning system, decision-tree merging system, and on-line predictive system. And we use the DTBMPA method to design the decision-tree merging system. Future directions of research are as follows. 1.Discussing the Decision-Tree preprocessing process in our Decision-Tree-Based Knowledge Integration Architecture. 2.Using the fuzzy theory to improve the accuracy of the merged tree when combining multiple predictions. 3.Discussing the merge of the complicated decision trees which are model trees, linear decision trees, oblique decision trees, regression trees, or fuzzy trees. 4.Discussing the process to merge two trees which have different possible values of non-numeric attributes or have different cut points of numeric attributes. 5.Comparing the performance of other pruning methods with ours. 6.Discussing the reconstruction of the merged trees after merging many new trees, discussing the adaptation of the merged trees to the changing environment, and observation of the evolution of the merged trees which are produced in different time stamp 7.Implementation of the on-line Decision-Tree-Based knowledge discovery in a real business environment.

The reliability and data rate of wireless communication have traditionally been limited by the presence of multipath fading in wireless channels. However, dramatic performance improvements can be obtained by the use of multiple transmit and receive antennas. Specifically, multiple antennas increase reliability by providing diversity gain, namely greater immunity to deep channel fades. They also increase data rates by providing multiplexing gain, i.e., the ability to multiplex multiple symbols in one signaling interval. Harvesting the potential benefits of multiple antennas requires the use of specially designed space-time codes at the transmitter front-end. Space-time codes introduce redundancy in the transmitted signal across two dimensions, namely multiple transmit antennas and multiple signaling intervals. In this work, we focus on linear space-time codes, which linearly combine the real and imaginary parts of their complex inputs to obtain transmit vectors for multiple signaling intervals. We aim to design optimum linear space-time codes. Optimality metrics and design principles for space-time codes are shown to depend strongly on the codes' function in the overall transmitter architecture. We consider two cases, depending on whether or not the space-time code is complemented by a powerful outer error-control code. In the absence of an outer code, the multiplexing gain of a space-time code is measured by its rate, while its diversity gain is measured by its raw diversity order. To maximize multiplexing and diversity gains, the space-time code must have maximum possible rate and raw diversity order. We show that there is an infinite set of maximum-rate codes, almost all of which also have maximum raw diversity order. However, different codes in this set have different error rate for a given input alphabet and SNR. Therefore, we develop analytical and numerical optimization techniques to find the code in this set which has the minimum union bound on error rate. Simulation results indicate that optimized codes yield significantly lower error rates than unoptimized codes, at the same data rate and SNR. In a concatenated architecture, a powerful outer code introduces redundancy in the space-time code inputs, obtaining additional diversity. Thus, the raw diversity order of the space-time inner code is only a lower limit to the total diversity order of the concatenated transmitter. On the other hand, we show that the rate of the space-time code places an upper limit on the multiplexing ability of the concatenated architecture. We conclude that space-time inner codes should have maximum possible rate but need not have high raw diversity order. In particular, the serial-to-parallel converter, which introduces no redundancy at all, is a near-optimum space-time inner code. This claim is supported by simulation results. On the receiver side, we generalize the well known sphere decoder to develop new detection algorithms for stand-alone space-time codes. These new algorithms are extended to obtain efficient soft-output decoding algorithms for space-time inner codes.

Thesis (MscAgric (Horticulture))--University of Stellenbosch, 2010. / ENGLISH ABSTRACT: Commercial fig production with popular European cultivars, Bourjasotte Noire, Col de Damme Noire and Noire de Caromb, is new to the Western Cape. Little research on fig production has been conducted in South Africa and producers are struggling to implement effective commercial practices. In order to establish practices that will maximise yield of quality fruit, the most productive one-year-old shoot lengths were identified in a phenological study. All shoot length categories evaluated in ‘Bourjasotte Noire’ (10 – 15 cm, 25 – 40 cm, 50 – 65 cm, 75+ cm) yielded fruit and will probably yield well the following season. In ‘Col de Damme Noire’, shoots longer than 60 cm seem to be suited to reproduction, yet they might produce a poor yield the following season. Shoots 10 – 20 cm long in ‘Noire de Caromb’ are productive relative to their length, while shoots 30 – 50 cm and 60 – 80 cm long are also fairly productive. Shoots longer than 100 cm produced suitable shoot lengths for yield the following season. Two experiments were conducted on ‘Bourjasotte Noire’ and ‘Col de Damme Noire’ to establish the type (Experiment 1, different intensity heading cuts) and timing (Experiment 2) of pruning cuts required to stimulate the growth of shoots of the same length as the shoots selected to be optimal for yield in the phenology study, and to reduce the expression of distal branching. In ‘Bourjasotte Noire’, removing one third of the total length of one-year-old shoots on 21 July by heading stimulated the development of more growth and longer current season shoots compared to other treatments, while reducing yield slightly. Heading back to three nodes in ‘Col de Damme Noire’ stimulated the growth of current season shoots of the optimal length established in the phenology study, while heading cuts on 30 June produced the longest average current season shoot length in ‘Col de Damme Noire’. To further address the effects of distal branching (acrotony), an experiment was conducted to establish whether rest breaking agents (RBA’s) in combination with tip-pruning can increase tree complexity by improving bud break, and whether increased complexity would increase yield in all three cultivars. A second experiment was conducted to evaluate the effects of timing RBA applications on bud break and harvest scheduling. Lift® increased the number of buds breaking in ‘Bourjasotte Noire’, while in ‘Noire de Caromb’ Dormex® and oil increased bud break. Tip-pruning increased the average shoot length in ‘Bourjasotte Noire’ and ‘Col de Damme Noire’, while causing a reduction in the amount of new growth in ‘Noire de Caromb’. Lift® applied 3 August and Dormex® applied 30 June shortened the number of days to 50% bud break in ‘Bourjasotte Noire’ and ‘Col de Damme Noire’. Both Lift® and Dormex® applied on 30 June decreased the number of days to 50% bud break and 50% harvest of the breba crop in ‘Noire de Caromb’. These treatments increased the number of fruit in both the breba and main crop of ‘Noire de Caromb’, but reduced fruit size. In conclusion, different approaches with regards to pruning needs to be followed for each cultivar to establish or maintain the optimal shoots for reproduction, while RBA’s can be used to force earlier, increased- bud break and harvest of breba fruit. / AFRIKAANSE OPSOMMING: Kommersiele verbouing van drie Europese vykultivars, Bourjasotte Noire, Col de Damme Noire en Noire de Caromb is nuut tot die Wes-Kaap. Baie min navorsing oor die verbouing van vye is al in Suid-Afrika gedoen, met die gevolg dat produsente sukkel om effektiewe kommersiële verbouingspraktyke te implementeer. ‘n Fenologiese studie van die drie kultivars is uitgevoer om vas te stel wat die mees produktiewe een-jaar-oue lootlengte is, met die doel om die opbrengs van kwaliteit vrugte te maksimeer. Al vier kategorieë wat ge-evalueer is in ‘Bourjasotte Noire’ (10 – 15 cm, 25 – 40 cm, 50 – 65 cm, 75+ cm) is gevind om geskik te wees vir huidige en volgende seisoen opbrengs. Lote langer as 60 cm is geskik vir opbrengs in ‘Col de Damme Noire’ in die huidige seisoen, maar mag in die volgende seisoen swak presteer a.g.v. die gebrekkige lengte van nuwe lote wat daarop ontwikkel. In ‘Noire de Caromb’ is gevind dat lote 10 – 20 cm lank baie produktief is relatief tot hul lengte en dat lote 20 – 50 cm en 60 – 80 cm lank ook relatief produktief is. Lote langer as 100 cm was minder produktief, maar het nuwe lote gelewer wat geskik is vir opbrengs die volgende seisoen. Twee snoei eskperimente is uitgevoer op ‘Bourjasotte Noire’ en ‘Col de Damme Noire’ om vas te stel wat die mees geskikte tipe snoeisnit (Eksperiment 1, verskillende dieptes van topsnitte) en tydstip om te snoei (Eksperiment 2) is met die doel om lote te produseer soortgelyk in lengte aan die wat in die fenologie studie uitgewys is as die produktiefste, en om moontlik die voorkoms van “kaalnekke” te verminder. In ‘Bourjasotte Noire’ is gevind dat die wegsnoei van ‘n derde van die loot op 21 Julie aanleiding gee tot meer groei, langer een-jaar-oue lote en ‘n effense afname in opbrengs. Geskikte lote langer as 60 cm kan in ‘Col de Damme Noire’ verkry word deur lote te top sodat net drie nodes oorbly. Die uitvoer van topsnitte op 30 Junie het langer gemiddelde lootlengtes tot gevolg gehad. Om die probleem van “kaalnekke” (apikale dominansie) verder aan te spreek, is ‘n eksperiment uitgevoer om vas te stel of rusbreekmiddels gekombineerd met tip-snoei gebruik kan word om kompleksiteit te vermeerder deur knopbreek te verhoog, en indien wel, of dit sal aanleiding gee tot verhoogde opbrengs in al drie kultivars. ‘n Tweede eksperiment met verskillende toedieningstye van rusbreekmiddels is uitgevoer om vas te stel of oeste geskeduleer kan word. Lift® het knopbreek verhoog in ‘Bourjasotte Noire’, terwyl Dormex® en olie knopbreek verhoog het in ‘Noire de Caromb’. Tip-snoei het die gemiddelde lootlengtes verhoog in ‘Bourjasotte Noire’ en ‘Col de Damme Noire’, terwyl dit groei verminder het in ‘Noire de Caromb’. Lift® toediening op 3 Augustus en Dormex® toediening op 30 Junie het die aanvang van 50% knopbreek vervroeg in ‘Boujasotte Noire’ en ‘Col de Damme Noire’. Beide Lift® en Dormex® toediening op 30 Junie het die bereiking van 50% knopbreek- en 50% oes vervroeg in ‘Noire de Caromb’. Hierdie behandeling het ook die aantal vrugte van die breba- en hoofoes vermeerder, maar vruggrootte verminder. Verskillende snoeibenaderings behoort dus gevolg te word om vir elke kultivar die regte lootlengtes te genereer of te onderhou, terwyl rusbreekmiddels gebruik kan word om knopbreek te vervroeg en verhoog, en om die breba oes te vervroeg.

Die Auswahl von Aktien ist ein Gebiet der Finanzanalyse, die von speziellem Interesse sowohl für viele professionelle Investoren als auch für Wissenschaftler ist. Empirische Untersuchungen belegen, dass Aktienerträge vorhergesagt werden können. Während verschiedene Modellierungstechniken zur Aktienselektion eingesetzt werden könnten, analysiert diese Arbeit die meist verbreiteten Methoden, darunter allgemeine Gleichgewichtsmodelle und Asset Pricing Modelle; parametrische, nichtparametrische und semiparametrische Regressionsmodelle; sowie beliebte Black-Box Klassifikationsmethoden. Aufgrund vorteilhafter Eigenschaften binärer Klassifikationsbäume, wie zum Beispiel einer herausragenden Interpretationsmöglichkeit von Entscheidungsregeln, wird der Kern des Handelsalgorithmus unter Verwendung dieser modernen, nichtparametrischen Methode konstruiert. Die optimale Größe des Baumes wird als der entscheidende Faktor für die Vorhersageperformance von Klassifikationsbäumen angesehen. Während eine Vielfalt alternativer populärer Bauminduktions- und Pruningtechniken existiert, die in dieser Studie kritisch gewürdigt werden, besteht eines der Hauptanliegen dieser Arbeit in einer neuartigen Methode asymmetrischen Baumprunings mit Abweisungsoption. Diese Methode wird als Best Node Selection (BNS) bezeichnet. Eine wichtige inverse Fortpflanzungseigenschaft der BNS wird bewiesen. Diese eröffnet eine einfache Möglichkeit, um die Suche der optimalen Baumgröße in der Praxis zu implementieren. Das traditionelle costcomplexity Pruning zeigt eine ähnliche Performance hinsichtlich der Baumgenauigkeit verglichen mit beliebten alternativen Techniken, und es stellt die Standard Pruningmethode für viele Anwendungen dar. Die BNS wird mit cost-complexity Pruning empirisch verglichen, indem zwei rekursive Portfolios aus DAX-Aktien zusammengestellt werden. Vorhersagen über die Performance für jede einzelne Aktie werden von Entscheidungsbäumen gemacht, die aktualisiert werden, sobald neue Marktinformationen erhältlich sind. Es wird gezeigt, dass die BNS der traditionellen Methode deutlich überlegen ist, und zwar sowohl gemäß den Backtesting Ergebnissen als auch nach dem Diebold-Marianto Test für statistische Signifikanz des Performanceunterschieds zwischen zwei Vorhersagemethoden. Ein weiteres neuartiges Charakteristikum dieser Arbeit liegt in der Verwendung individueller Entscheidungsregeln für jede einzelne Aktie im Unterschied zum traditionellen Zusammenfassen lernender Muster. Empirische Daten in Form individueller Entscheidungsregeln für einen zufällig ausgesuchten Zeitpunkt in der Überprüfungsreihe rechtfertigen diese Methode. / Stock picking is the field of financial analysis that is of particular interest for many professional investors and researchers. There is a lot of research evidence supporting the fact that stock returns can effectively be forecasted. While various modeling techniques could be employed for stock price prediction, a critical analysis of popular methods including general equilibrium and asset pricing models; parametric, non- and semiparametric regression models; and popular black box classification approaches is provided. Due to advantageous properties of binary classification trees including excellent level of interpretability of decision rules, the trading algorithm core is built employing this modern nonparametric method. Optimal tree size is believed to be the crucial factor of forecasting performance of classification trees. While there exists a set of widely adopted alternative tree induction and pruning techniques, which are critically examined in the study, one of the main contributions of this work is a novel methodology of nonsymmetrical tree pruning with reject option called Best Node Selection (BNS). An important inverse propagation property of BNS is proven that provides an easy way to implement the search for the optimal tree size in practice. Traditional cost-complexity pruning shows similar performance in terms of tree accuracy when assessed against popular alternative techniques, and it is the default pruning method for many applications. BNS is compared with costcomplexity pruning empirically by composing two recursive portfolios out of DAX30 stocks. Performance forecasts for each of the stocks are provided by constructed decision trees that are updated when new market information becomes available. It is shown that BNS clearly outperforms the traditional approach according to the backtesting results and the Diebold-Mariano test for statistical significance of the performance difference between two forecasting methods. Another novel feature of this work is the use of individual decision rules for each stock as opposed to pooling of learning samples, which is done traditionally. Empirical data in the form of provided individual decision rules for a randomly selected time point in the backtesting set justify this approach.