No description available.

No description available.

No description available.

No description available.

No description available.

An asset manager's goal is to provide a high return relative the risk taken, and thus faces the challenge of how to choose an optimal portfolio. Many mathematical methods have been developed to achieve a good balance between these attributes and using di erent risk measures. In thisthesis, we test the use of a relatively simple and common approach: the Markowitz mean-variance method, and a more quantitatively demanding approach: the tail optimization method. Using active portfolio based on data provided by the Swedish fund management company Enter Fonderwe implement these approaches and compare the results. We analyze how each method weighs theunderlying assets in order to get an optimal portfolio.

A map combines scientific facts with aesthetic perceptions. This study argues that scaling is universal in mapping reality and evoking a sense of beauty. Scaling laws are used to reveal the underlying structures and dynamics of spatial features. Complex systems, such as living cities involve various interacting entities at all scales. Each individual coherently interacts and overlaps with others to create an unbreakable entity. Scaling structures are also known as fractals. Fractal geometry is used to depict a complex system. Natural objects, such as trees, contain a similar geometry (branches) at all scales. This study attempts to effectively visualize the scaling pattern of geographic space. In this regard, the head/tail breaks classification is applied to visualize the scaling pattern of spatial features. A scaling pattern underlies a geographic space. Visualizing the scaling structure using the head/tail breaks classification can further evoke a sense of beauty. This kind of beauty is on the structural level and was identified by Christopher Alexander, who asserted that beauty is not a personal experience but objectively exists in any space. Alexander developed the theory of centers to broaden the concepts of life and beauty.  A structure with a scaling property (with recursive centers) has high quality of life, and a scaling pattern has positive effects on individual’s psychological and physical well-being. To verify the concept of objective beauty, human beings are used as measuring instruments to examine the assumptions. This study adopts the mirror-of-the-self test to examine human reactions to 23 pairs of images, including photographs of buildings and two types of map. The idea is that participants sense the quality of life by comparing a pair of objects and selecting the object that presents a better picture of themselves. Once individuals feel the self in a picture, they are able to detect real beauty. In this manner, individuals can detect real beauty and life that deeply connect to their inner hearts. The tests were conducted through personal interviews and Internet surveys with the public and with professionals, and 392 samples were collected. The study results show that more than 60% of the individuals selected images with a scaling pattern. These results are in accordance with Alexander’s assumption. In particular, more than 65% individuals selected maps that depict scaling forms. Moreover, this study conducted a training test with a particular group of individuals, after which more than 70% of individuals selected scaling maps. The results reveal that scaling laws are applicable for creating maps and evoking a sense of beauty.

Vitamin C (ascorbic acid, AsA), an important primary metabolite of plants, functions as an antioxidant, an enzyme cofactor, and a cell-signaling modulator in a wide array of crucial physiological processes including biosynthesis of the cell wall, secondary metabolites and phytohormones, stress resistance, photoprotection, cell division, senescence, and growth. To identify genes that may regulate vitamin C levels in plants, about 3000 activation-tagged Arabidopsis lines were treated with ozone, which is a power oxidizing agent. Two mutants were selected for identification of potential genes involved in the regulation of vitamin C synthesis. A putative F-box gene, VCF1, and a purple acid phosphatase, AtPAP15, were identified for further characterization. Two homozygous SALK T-DNA knockouts in the open reading frame (ORF) of VCF1 exhibited high tolerance to ozone when treated with 450 ppb for 3 hours and the AsA levels of these mutants were 2 to 3 fold higher than wild-type (wt) plants. Developmental studies, using RT-PCR, indicated that foliar expression of the VCF1 gene increased with plant age from 1 to 5 weeks, whereas AsA decreased during this same period. The expression of VCF1 was higher under a low-light condition in which AsA was reduced considerably. The AsA levels in two VCF1 overexpressing lines were only 50 to 70% of wt plants. These results suggested that the putative F-box gene functions as a negative regulator of leaf ascorbate content. Overexpression of AtPAP15 with the CaMV 35S promoter resulted in up to 3-fold higher AsA levels than wt plants, where two independent SALK T-DNA insertion mutants in AtPAP15 had 50% less AsA than wt plants. Enzyme activity of bacterially expressed GST:AtPAP15 was greatest with phytate as a substrate indicating that AtPAP15 is a phytase. Phytase catalyzes hydrolysis of phytate (myo-inositol hexakisphosphate) to yield myo-inositol and free phosphate. Thus, AtPAP15 may regulate AsA levels by controlling the input of myo-inositol into this branch of AsA biosynthesis in Arabidopsis. AtPAP15 was expressed in all tested organs in wt plants and suggests that the enzyme may have functions other than phytate degradation during seed germination. / Ph. D.

Tails are frequently utilized by animals to enhance their motion agility, dexterity, and versatility, such as a cheetah using its tail to change its body orientation while its legs are all off the ground and a monkey using its tail to stabilize its locomotion on branches. However, limited by technology and application scenarios, most existing legged robots do not include a robotic tail on board. This research aims to explore the possibilities of adding this missing part on legged robots and investigate the tail's functionalities on enhancing the agility, dexterity, and versatility of legged locomotion. In particular, this research focuses on animal-like serpentine tail structure, due to its larger workspace and higher dexterity. The overall research approach consists of two branches: a theoretical branch that focuses on dynamic modeling, analysis, and control of the legged robots with a serpentine robotic tail; and an empirical branch that focuses on hardware development and experiments of novel serpentine robotic tails and novel legged robots with tail. More specifically, the theoretical work includes modeling and control of a general quadruped platform and a general biped platform, equipped with one of the two general serpentine tail structures: an articulated-structure tail or a continuum-structure tail. Virtual work principle-based formulation was used to formulate the dynamic model. Both classic feedback linearization-based control and optimization-based control were used to coordinate the leg motions and the tail motion. Comparative studies on different tail structures as well as numerical analyses on robotic locomotion were performed to investigate the dynamic effects of serpentine robotic tails. The empirical work includes the developments and experiments of two novel serpentine robotic tail mechanisms and one first-of-its-kind quadruped robot ("VT Lemur") equipped with a serpentine robotic tail. To develop these novel robots, a systematic approach based on dynamic analysis was used. Various experiments were then conducted using the robot hardware. Both the theoretical and empirical results showed that the serpentine robotic tail has significant effects on enhancing the agility, dexterity, and versatility of legged robot motion. / Doctor of Philosophy / Quadruped robots have made impressive progresses over the past decade and now can easily achieve complicated, highly dynamic motions, such as the backflip of the MIT Mini Cheetah robot and the gymnastic parkour motions of the Atlas robot from Boston Dynamics, Inc. However, by looking at nature, many animals use tails to achieve highly agile and dexterous motions. For instance, monkeys are observed to use their tails to grasp branches and to balance their bodies during walking. Kangaroos are found to use their tails as additional limbs to propel and assist their locomotion. Cheetahs and kangaroo rats are thought to use their tails to help maneuvering. Therefore, this research aims to understand the fundamental principles behind these biological observations and develop novel legged robots equipped with a serpentine robotic tail. More specifically, this research aims to answer three key questions: (1) what are the functional benefits of adding a serpentine robotic tail to assist legged locomotion, (2) how do animals control their tail motion, and (3) how could we learn from these findings and enhance the agility, dexterity, and versatility of existing legged robots. To answer these questions, both theoretical investigations and experimental hardware testing were performed. The theoretical work establishes general dynamic models of legged robots with either an articulated tail or a continuum tail. A corresponding motion control framework was also developed to coordinate the leg and tail motions. To verify the proposed theoretical framework, a novel quadruped robot with a serpentine robotic tail was developed and tested.

The rapid advancement of Machine Learning (ML) and Deep Learning (DL) has revolutionized various domains, necessitating efficient and cost-effective ML inference capabilities. Function-as-a-Service (FaaS) has emerged as a promising approach for hosting ML inference services, providing a serverless computing environment that streamlines development cycles and offers scalability and simplified infrastructure management. However, existing autoscaling strategies employed by popular FaaS platforms often overlook critical factors such as response time and tail latency. Additionally, Python's Global Interpreter Lock (GIL) poses challenges for parallel computing in high-request traffic scenarios. This thesis addresses the need for efficient and cost-effective Machine Learning (ML) inference capabilities by exploring batching and autoscaling strategies for Serverless Inference instances. The study proposes a prototype FaaS framework that provides adaptive request batching, reactive autoscaling policies, and SLO monitoring, thus allowing Serverless Inference workloads to meet their SLO targets even during peak traffic. The proposed approach aims to optimize resource utilization, mitigate tail latency, and improve overall system performance. / Master of Science / Machine Learning (ML) and Deep Learning (DL) are advanced techniques that allow computers to learn from data and make predictions or decisions without being explicitly programmed. This has led to significant advancements in various fields. Inference refers to the process of applying a trained ML model to new data to make predictions or extract insights. In the context of ML, there is a growing need for efficient and cost-effective inference capabilities. A new approach called Function-as-a-Service (FaaS) has emerged that can address this need. FaaS is a way of abstracting the server infrastructure away from the developers. This means developers can focus on writing the ML code without worrying about managing the underlying infrastructure. FaaS offers benefits such as scalability, simplified infrastructure management, and faster development cycles. However, existing FaaS platforms face challenges in ensuring fast response times and handling high levels of incoming requests. This thesis aims to address these challenges by proposing a prototype FaaS framework. The framework incorporates adaptive request batching, reactive autoscaling policies, and Service-Level Objectives (SLOs) monitoring. Request batching allows the framework to process multiple requests together, improving efficiency. Autoscaling policies ensure the system dynamically adjusts its resources based on the incoming workload. Monitoring SLOs helps track and meet performance targets, even during peak traffic. By optimizing resource utilization, reducing delays in processing requests, and improving overall system performance, the proposed approach seeks to provide efficient and cost-effective ML inference capabilities in a serverless environment.