Across the academic sphere, much research has been conducted into the development of water-sensitive elements to address issues around urban water management. However, these elements are commonly investigated in isolation, with little consideration for initiatives from other disciplines that may support their success. This research aims to demonstrate the value that an architect may bring in incorporating ideas drawn from various disciplines to create a water- sensitive design solution with multiple ecosystem benefits, taking into account the human experience of space and place-making. In doing so, the design demonstrates that a water-sensitive building is aesthetically pleasing, viable and achievable. The feasibility of water-sensitive designs has been noted as a focus area by the South African Water Research Commission; one which is particularly pertinent in our present water-scarce environment in South Africa. This applied study is based on a previous Master of Architecture (Professional) dissertation building design, which is used as the unit of analysis. The building focuses on restoring the quality of water in the Liesbeek River in Cape Town using passive filtration methods. The objective of this study is to gain new insights into the design process and planning of water-sensitive architectural buildings, which assists in understanding when collaborating across disciplines. The research is guided by Deep Ecology, phenomenology and Ecological Urbanism. Research by Design is used as the method of the study, in which different design iterations based on the raw data of the original building are investigated and analysed, as well as evaluated by specialists from various disciplines in order to create a best-fit design solution. The revised building takes into account the practical, site-specific and architectural qualities of a water-sensitive design to create a people-centred building that incorporates ecological and engineering demands in greater detail. Key outcomes of the study include a typical design process for a WSAD and architectural guidelines for water-sensitive buildings, grounded in the diverse values of water and its relationship to people and nature. The dissertation aims to contribute to the academic discourse around water-sensitive design. Further, the guidelines developed may be used to inform the design of conventional buildings.

Earthquake detection is an important task, focusing on detecting seismic events in past data or in real time from seismic time series. In the past few decades, due to the increasing amount of available seismic data, research in seismic event detection shows remarkable success using neural networks and other machine learning techniques. However, creating high quality labeled data sets is still a manual process that demands tremendous amount of time and expert knowledge, and is stifling big data innovation. When compiling a data set, it is unclear how many earthquakes and noise are mislabeled. Another challenge is how to promote the general applicability of the machine learning based models to different geographical regions. The models trained by data sets from one location should be applicable to the detection at other locations. This thesis explores the most popular deep learning model, convolutional neural networks (CNN), to build a single location detection model. In addition, we build more robust generalized earthquake detection models using transfer learning and meta learning. We also introduce a process for generating high quality labeled datasets. Our technique achieves high detection accuracy even on low signal to noise ratio events. The AI techniques explored in this research have potential to be transferred to other domains that utilize signal processing. There are a myriad of potential applications, with audio processing probably being one of the most directly relevant. Any field that deals with waveforms (e.g. seismic, audio, light) can utilize the developed techniques.

The amount of available protein sequences is rapidly increasing, mainly as a consequence of the development and application of high throughput sequencing technologies in the life sciences. It is a key question in the life sciences to identify the functions of proteins, and furthermore to identify the phenotypes that may be associated with a loss (or gain) of function in these proteins. Protein functions are generally determined experimentally, and it is clear that experimental determination of protein functions will not scale to the current { and rapidly increasing { amount of available protein sequences (over 300 million). Furthermore, identifying phenotypes resulting from loss of function is even more challenging as the phenotype is modi ed by whole organism interactions and environmental variables. It is clear that accurate computational prediction of protein functions and loss of function phenotypes would be of signi cant value both to academic research and to the biotechnology industry. We developed and expanded novel methods for representation learning, predicting protein functions and their loss of function phenotypes. We use deep neural network algorithm and combine them with symbolic inference into neural-symbolic algorithms. Our work signi cantly improves previously developed methods for predicting protein functions through methodological advances in machine learning, incorporation of broader data types that may be predictive of functions, and improved systems for neural-symbolic integration. The methods we developed are generic and can be applied to other domains in which similar types of structured and unstructured information exist. In future, our methods can be applied to prediction of protein function for metagenomic samples in order to evaluate the potential for discovery of novel proteins of industrial value. Also our methods can be applied to the prediction of loss of function phenotypes in human genetics and incorporate the results in a variant prioritization tool that can be applied to diagnose patients with Mendelian disorders.

Recognizing visual information is crucial for many real artificial-intelligence-based applications, ranging from domestic robots to autonomous vehicles. However, the success of deep learning methods on visual recognition tasks is highly dependent on access to large-scale labeled datasets, which are expensive and cumbersome to collect. Transfer learning provides a way to alleviate the burden of annotating data, which transfers the knowledge learned from a rich-labeled source domain to a scarce-labeled target domain. However, the performance of deep learning models degrades significantly when testing on novel domains due to the presence of domain shift. To tackle the domain shift, conventional domain adaptation methods diminish the domain shift between two domains with a distribution matching loss or adversarial loss. These models align the domain-specific feature distribution and the domain-invariant feature distribution simultaneously, which is sub-optimal towards solving deep domain adaptation tasks, given that deep neural networks are known to extract features in which multiple hidden factors are highly entangled. This thesis explores how to learn effective transferable features by disentangling the deep features. The following questions are studied: (1) how to disentangle the deep features into domain-invariant and domain-specific features? (2) how would feature disentanglement help to learn transferable features under a synthetic-to-real domain adaptation scenario? (3) how would feature disentanglement facilitate transfer learning with multiple source or target domains? (4) how to leverage feature disentanglement to boost the performance in a federated system? To address these needs, this thesis proposes deep adversarial feature disentanglement: a class/domain identifier is trained on the labeled source domain and the disentangler generates features to fool the class/domain identifier. Extensive experiments and empirical analysis demonstrate the effectiveness of the feature disentanglement method on many real-world domain adaptation tasks. Specifically, the following three unsupervised domain adaptation scenarios are explored: (1) domain agnostic learning with disentangled representations, (2) unsupervised federated domain adaptation, (3) multi-source domain adaptation.

京都大学 / 新制・課程博士 / 博士(工学) / 甲第23173号 / 工博第4817号 / 新制||工||1753(附属図書館) / 京都大学大学院工学研究科都市社会工学専攻 / (主査)教授 堀 智晴, 教授 田中 茂信, 教授 角 哲也 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DFAM

Up to 87% of the world’s wetlands have been destroyed, considerably reducing ecosystem services these wetlands once provided. More recently, many wetlands are being restored in an attempt to regain their ecosystem service. This study seeks to determine the effects of restored wetlands on local bat habitat use. Bat activity was found to be significantly higher around the wetlands when compared to distant grassy fields; however, no significant difference was found among the restored wetlands and a remote cattle farm containing multiple water features. Geospatial models of bat distribution and bat foraging were produced using machine learning that showed higher habitat suitability and foraging activity around restored wetlands than around distant grassy fields, suggesting that wetlands provide vital habitat for insectivorous bats. This study demonstrates that restored wetlands promote bat activity and bat foraging, and restoring wetlands may be a useful means of increasing natural pest control over nearby farmlands.

Stemming from the growth and increased complexity of computer vision, natural language processing, and speech recognition algorithms; the need for scalability and fault tolerance of machine learning systems has risen. In order to comply with these demands many have turned their focus towards implementing machine learning on distributed systems. When running time demanding and resource intensive tasks like machine learning training on a cluster, resource efficiency is very important to keep training time low. To achieve efficient resource allocation a cluster scheduler is used. Standard scheduling frameworks are however not designed for deep learning, due to their static resource allocation. Most frameworks also do not make use of a serverless architecture, despite its ease of management and rapid scalability making it a fitting choice for deep learning tasks. Therefore we present Coach, a serverless job scheduler specialized for parameter server based deep learning models. Coach makes decisions to maximize resource efficiency and minimize training time through use of regression techniques to fit functions to data from previous training epochs. With Coach we attempt to answer three questions concerning the training speed (epochs/second) of deep learning models on a distributed system when using a serverless architecture. The three questions are as follows. One: does the addition of more workers and parameter servers have a positive impact on the training speed when running a varying number of concurrent training jobs? Two: can we see improved performance in regards to the training speed, when training is done in a distributed manner on a cluster with limited resources, compared to when it is done on a singular node? Three: how accurate are predictions made using fitted functions of previous training data at estimating the optimal number of workers and parameter servers to use during training, in order to maximize training speed? Due to limitations with the cluster used for testing we see that a minimal setup of a singular worker and server is almost always optimal. With results indicating that an additional server can have slight positive effects in some situations and an additional worker only appears positive in high variance situation where there are many jobs running at the same time. Which is theorized to be caused by choices made by the Kubernetes scheduler.

Designing games around complex emotions and experiences can be seen as quite difficult for game designers. Therefore, this paper sought to research how to convey the emotional state of an individual through design based on a chosen experience. The chosen experience was the feeling of loneliness with connection to the fear of rejection. This paper will be using Rusch’s (2017) method of designing deep games as a way of modeling an experience of loneliness as perceived by one of the authors. This model will be based on a conversation between the two authors with one acting as an experience expert and the other one as a listener. The results from this method seek to gain a catalog of experiences and thoughts about the experience expert’s feelings of loneliness that can be used to construct metaphors. These metaphors are then used to create game mechanics that combined hope to achieve a sincere recreation of the experience expert feelings of loneliness. This paper will go into depth on the design decisions made during the production of the deep game modeling the emotion loneliness. / Att designa spel kring komplexa känslor och upplevelser kan ses som en svårighet för speldesigners. Därför försökte denna uppsats undersöka hur man förmedlar en individs emotionella tillstånd genom design baserat på en vald upplevelse. Den valda upplevelsen var känslan av ensamhet med koppling till rädslan av förkastelse. Denna uppsats kommer att använda Ruschs (2017) metod för att designa djupa spel som ett sätt för att modellera en upplevelse av ensamhet som uppfattas av en av författarna. Denna modell kommer att baseras på en intervju mellan de två författarna med en som tar rollen av en erfarenhetsexpert och den andra som en lyssnare. Resultaten från denna metod försöker få en katalog med upplevelser och tankar om erfarenhetsexpertens känslor av ensamhet som kan användas för att konstruera metaforer. Dessa metaforer används sedan för att skapa spelmekaniker som kan kombineras med hopp om att uppnå en uppriktig rekreation av den individuella känslan av ensamhet som erfarenhetsexperten uplever. Denna uppsats kommer att gå djupare in på de designbeslut som fattas under produktionen av det djupa spelet som modellerar känslan av ensamhet.

No description available.

The naturally occurring isotopes 231Pa and 230Th are used as tracers of marine biogeochemical processes. They are both produced from the radioactive decay of their uniformly distributed uranium parents (235U and 234U) in seawater. After production, 231Pa and 230Th are removed by adsorption onto settling particles (scavenging) and subsequently buried in marine sediments. 230Th is more particle reactive compared to 231Pa. Consequently, 230Th will be removed from the open ocean by adsorption onto settling particles, while 231Pa tend to be laterally transported by currents and removed by scavenging in areas of high particle flux (e.g. ocean margins). The primordial 232Th indicates lithogenic supply via rivers and resuspension of sediments, which provides additional information about processes involved in the cycling of particle reactive elements in the ocean. The preferential deposition of particle reactive elements at ocean margins (boundary scavenging) has important implications for our understanding of the distribution and dispersion of micronutrients (e.g. iron) and pollutants in the ocean. It is therefore valuable to understand the nature of boundary scavenging processes in order to evaluate the relative contribution of circulation and scavenging behaviors.The major characteristics of thermohaline circulation in the Mediterranean are well known and have been studied for decades. This sea is an almost land-locked area, where limited water-exchange with the Atlantic Ocean only occurs through the Strait of Gibraltar. Therefore, this marginal sea is often referred to as a “miniature ocean” suitable as a “laboratory” for marine environmental research. In this licentiate thesis, distributions of 231Pa, 230Th and 232Th in seawater and marine particles collected during the GEOTRACES MedSeA-GA04-S cruise in 2013 are presented. Observed nuclide distributions indicate the impact of deep water formation processes, where observed differences can be linked to the type of deep water formation process that occurs in respective basin. Essentially all in-situ produced 230Th is buried in Mediterranean Sea sediments. Despite lower affinity of 231Pa for marine particles, most 231Pa is also scavenged and deposited in Mediterranean Sea sediments. The efficient scavenging of 231Pa produces a relatively low fractionation between 231Pa and 230Th in terms of the fractionation factor FTh/Pa. This licentiate thesis presents a summary of the methods used for the analysis of 231Pa and Th-isotopes with details on the exchange chromatography method and the treatment of mass spectrometric data. The study of 231Pa, 230Th and 232Th in the Mediterranean Sea has important implications for our understanding of processes that control their water column distributions and how their behavior can be utilized to trace chemical flux in modern and past ocean environments. / GEOTRACES / MeDSeA