Refine
Document Type
- Bachelor thesis (5) (remove)
Has Fulltext
- yes (5)
Is part of the Bibliography
- no (5)
Language
- English (5) (remove)
Keywords
- Sustainability (2)
- AMPL (1)
- Abwanderungsvorhersage (1)
- Artificial neuronal networks (1)
- Atmung (1)
- Bilderkennung (1)
- Bodenatmung (1)
- Churn prediction (1)
- Critical theory (1)
- Environment (1)
Within the last decades, the number of social networks is growing fast. The competition of retaining the customers to grow their platform and increase their profitability is rising. That is why companies need to detect possible churners to retain these. The problem of predicting the users’ lifetime, churning users, and the reasons for churning can be tackled by using machine learning.
The goal of this bachelor thesis is to build machine learning models to predict user churn and the user lifetime within the social network Jodel, a location-based anonymous messaging application for Android and iOS.
To get the best possible prediction results, we have started with extensive literature research, whose approaches we have tested and added to a machine learning pipeline to build predictive models. With these models, we have investigated the performance after different observation time windows and have finally compared the strongest models to detect similarities and understand the insights to learn their behaviour.
The results of this thesis are machine learning models for a selected representative set of communities varying in size within the Kingdom of Saudi Arabia and a country model leveraging all data. These models are used for a regression task by predicting the lifetime of a user and a multi-label classification of a user into six different churn classes. Additionally, we have also given models for a binary classification, where the model will predict if the user will churn within a given time or not. These models have shown general strong predictive power, which is shrinking when limiting the observation time window. Especially the binary classification yielded high accuracy of over 99%.
The best models have been used for predicting user churn within other communities to detect communities with possible similar behaviour. These similarities then have been determined by features’ importance, where the most important features have got fed back into empirics. This has shown statistically significant differences between user groups with a different active time but as of today no clear trends were visible that had led us to define the communities’ behaviours.
Since the competition of social networks is still growing, the retaining of users will stay a core marketing strategy, which will need to be tackled by machine learning and artificial intelligence. The created models could be useful for predicting churning users within the platform Jodel to detect these customers that will churn within a given time.
Researches did not focus much on anonymous and location-based messaging. That is why the results of this thesis on the anonymous messaging application Jodel opens a variety of possible tasks for the future in this context.
Machine learning is a field that has been the object of study of many researchers around the globe during the last decades. Very often to solve machine learning challenges like classification problems for example, one needs to train an artificial neural network. To train this network a certain loss function has to be minimized. There is a ubiquitous approach to achieve this which consists of using variants of the stochastic gradient descent combined with the backpropagation algorithm. In our work, we aimed at testing a rather non-conventional scheme consisting of making use of the solvers a software called AMPL offers.
Flooding of dry sediments is known to trigger pulses of microbial respiration at land-water-interfaces. The regulation of discharge variability is therefore proposed to affect the respiration balance of these sites. In this study, I assessed the impact of discharge regulation on microbial respiration associated to surface sediments at a land-water-interface of the river Spree. I developed a theoretical model, based on empirical respiration data, to estimate the two-month total respiration at the study site for three discharge scenarios. The real scenario represented the actual discharge at the study site, which was regulated by the Spremberg reservoir dam. In the unregulated scenario, the regulating effect by the dam was excluded. In the extremely regulated scenario, a hypothetical constant discharge was modeled. For each scenario, the daily discharge, the corresponding flooded areas, the extent of dry or rewetted areas, and the durations of flooding or rewetting by rain at the study site were determined. Microbial respiration rates associated to surface sediments were measured with a respirometer under flooded, dry, and rewetted conditions. The model applied these rates to the respective flooded, dry or rewetted areas of the study site, to calculate the daily areal respiration. In all sediments from the land-water-interface, a distinctive respiration pulse was measured on the first day of flooding, and higher respiration rates under flooded than under dry conditions. The discharge of the unregulated scenario was characterized by a higher variability and larger flow volume than the regulated real scenario. Due to the higher total discharge, larger areas were flooded in the unregulated scenario, the total respiration from sediments under long-term flooded conditions was therefore higher. Moreover, the daily extent of flooded areas fluctuated more strongly, hence more short-term respiration pulses upon flooding were triggered than in the real regulated scenario. The calculated total two-month respiration of the unregulated scenario exceeded that of the regulated real scenario by almost 14 %. These results suggest that discharge regulation can have a considerable negative impact on sediment-associated microbial respiration at land-water-interfaces.
Conflicts emerging through large-scale mining projects among government, industry and affected population are common all over the world. Especially the indigenous communities and their strong human-nature relation are threatened by the land transformation due to mining operations. Open pit gold mining performed by transnational companies TNCs and supported by the Philippine government aligning decisions on economic standards and market orientated legislative period, have been imposing tremendous damage on the Philippine indigenous communities and the national environmental resources for many decades, from colonial times, dictatorship, and new democracy . This thesis displays and discusses the interwoven web of socially and environmentally degrading impacts introduced to the small village of Didipio in the north of the Philippines by a planned gold and copper mining project. After describing influencing sectors of industry and government for decision making in mining projects and introducing interests and purposes of all stakeholders, assessment of changes in water quality and quantity, which are associated with technical requirements of open pit gold mining, is used to specifically relate the initiation of a punctual impact, on a natural good, and its provoked tremendous environmental and furthermore social degradation. A central assertion of this thesis is the call for recognizing human-rights based participation of all affected stakeholders via emphasizing indigenous people’s rights, balancing market-centred development aggression and foreign control over national markets, as well as implementing sustainable environmental awareness management by industry and government. Further recommendations are related to the resignation of defining nature merely according to its function under the utilization concept, to TNCs’ application of postmodern understanding of reason based on flexibility and versatility, and to the promoting of indigenous self-determination and individual initiative.
Map-Making for ERM Studies
(2005)
This paper discusses the need for a concept “environmental and resource management” (ERM), and what ERM meaningfully could be. This problem is approached first by an analysis of the environmental and developmental issues. Consequently the hegemonic discourse on sustainability is questioned. Second, the paper uses Kuhn’s “Structure of Scientific Revolutions” as a basis for drafting a scientific field “ERM”. From the perspective of both approaches, different paradigms of environmental management are studied. Based on this analysis a fundamental problem is obtained, namely, how to make individuals and institutions co-operate and under which societal conditions co-operation can be engendered. Subsequently ERM is defined. ERM-Science is being worked out, as being goal-driven, seeking sustainability. It is pointed out that ERM, ERM-Science and its supporting scientific fields are not identical. The major findings following this method are: ERM is stipulated as an inherent social activity. Therefore the scientific field ERM needs to critically focus on the environmental and resource (ER)-manager’s behaviour and her context, that both have indicated and non-indicated environmental influences upon the world. To inquire in that object of interest, the following point of view is developed: “How to make people and society not only speak about environmental problems but implement solutions?” This leads to the discussion of consequences in terms of ethical implications and methodology. Basing on these results, the legitimate problems for the scientific field ERM are illustrated. The developed definition and its consequences are applied on the three ERM-programs at Brandenburg University of Technology (BTU) (Germany) and concludes that a radical reform of the BTU-approach to ERM is required.