MLOps and Kubernetes are two major components of the modern-day information technology landscape, and their impact on the field is likely to grow even stronger in the near future. As a multitude of tools have been developed for the purpose of facilitating effortless creation of cloud native MLOps solutions, many of them have been designed, to varying degrees, to integrate with the Kubernetes system. While numerous evaluations have been conducted on these tools from a general MLOps perspective, this thesis aims to evaluate their qualities specifically within a Kubernetes context, with the focus being on their integration into this ecosystem. The evaluation is conducted in two steps: an MLOps market overview study, as well as an in-depth MLOps tool evaluation. The former represents a macroscopic overview of currently available MLOps tooling, whereas the latter delves into the practical aspects of deploying three Kubernetes native, open source MLOps platforms on cloud-based Kubernetes clusters. The platforms are Kubeflow, Pachyderm, and Polyaxon, and these are evaluated in terms of functionality, usability, vitality, and performance.

Syftet med denna avhandling var att undersöka och utvärdera vilka val utvecklare bör göra vid skapandet av mobila applikationer. Vidare var målet att hitta lämpliga ersättare för Android delen i kursen HI1033 Mobila applikationer och trådlösa nät på Kungliga Tekniska Högskolan. För att åstadkomma detta utvärderades de mest populära plattformsoberoenderamverken på marknaden, Flutter och React, tillsammans med traditionella och nya ’native’ alternativ som Jetpack Compose. Granskningen av mobilapplikationerna inkluderade utvärderingar av utseende, tidsåtgång, komplexitet, dokumentation, stabilitet, kodlängd samt genomförbarhet bland många. Applikationerna utvecklades med dessa i åtanke, men krävdes också att följa de obligatoriska kraven för mobila applikationskursen. Detta för att ge svar på om dessa varianter kunde anpassas i kursen för framtida klasser. Om applikationerna inte kunde följa kursens krav var tillvägagångsättet sannolikt inte ett genomförbart alternativ. Resultaten visade att, tillskillnad från Jetpack Compose är Flutter betydligt svårare och mer komplicerat att implementera. Det har potential, men det saknar komplex funktionalitet till Blåtand-sensorer och mer specifika ’native’ funktioner.  I jämförelsen med Jetpack Compose och traditionell Android utveckling i denna fråga visades det att Jetpack Compose var både enklare och mindre tidskrävande, kortare i kodlängd men även kunde replikeras på utseendet. Därför är Jetpack Compose ett bra alternativ för att bygga ’native’ applikationer samtidigt som det uppfyller kraven för HI1033 Mobila applikationer och trådlösa nät. / The purpose of this thesis was to study and evaluate what choices developers must make when creating mobile applications. Further, the aim is to find a suitable replacement for the Android part of the course in HI1033 Mobile Applications and Wireless Networks at the KTH Royal institute of Technology. To accomplish this, the most popular cross-platform frameworks on the market, Flutter and React, along with traditional and new native alternatives such as Jetpack Compose, were evaluated.  The examination of the mobile applications included evaluations of appearance, time consumption, complexity, stability, and feasibility among many. The applications were developed with this in mind, however, they needed to also follow the mandatory requirements of the mobile application course to provide answers as to whether these variants could be adapted in the course’s future classes. If the applications could not meet the course's requirements, then the approach was likely not a viable one. The results indicated that, unlike Jetpack Compose, Flutter is significantly more difficult and cumbersome to implement. It has potential, but it’s missing complex functionality for Bluetooth sensors and more specifically certain native functionalities. Comparing Jetpack Compose with traditional android development in this question proved that the newer Jetpack Compose was both easier, less time-consuming, and more stable to use. Therefore, Jetpack Compose is a good alternative for building native mobile applications, but also for the future classes in HI1033 Mobile Application and Wireless Networks.

Interseeding legumes in native warm-season grasses (NWSG) may improve the nutritive value of the stand, result in more consistent forage availability throughout the growing season, and increase forage yield. These benefits are often not realized due to difficulties in establishing legumes in existing NWSG stands. The objective of this study was to investigate the effect of planting method of legume interseeding, timing of legume interseeding, and the efficacy of burning plant residue on legume establishment in NWSG. Two forage legumes, 'Alice' white clover (Trifolium repens L.) and 'Freedom HR' red clover (Trifolium pratense L.), were interseeded into mixed 'Niagara' big bluestem (Andropogon gerardii Vitman), 'GA Ecotype' Indiangrass (Sorghastrum nutans Nash), and 'Camper' little bluestem (Schizachyrium scoparium) pasture in 2022 and 2023 at the Southern Piedmont AREC in Blackstone, Virginia. Planting method at three levels (no-till drill, broadcast, and non-planted control) were evaluated at three planting timing levels (fall planting, winter planting, and winter planting with burned residue). Among the treatment combinations, burned plots that were drilled resulted with the greatest spring clover count of 236 plants m-2, followed by winter drill (146 plants m-2) and burn broadcast (133 plants m-2). All fall plantings and all control plots were similar with a mean of 21 plants m-2. As a result of greater initial clover emergence, plots that were burned or seeded in the winter had greater clover content throughout the experiment; burned and drilled plots had over 90% clover ground cover throughout the second year. Domination of plots by clover in the second year caused yields and the proportion of NWSG in the stand to decline, with burned plots yielding 5,757 kg ha-1 compared to a winter-fall mean of 7,429 kg ha-1. Plots with greater clover content were able to sustain higher crude protein content and lower neutral detergent fiber content in both the establishment year and the second year. Though interseeding legumes benefitted nutritive values, these results suggest that red clover may be incompatible with the NWSG evaluated. Burned plots were especially affected by excessive competition. Further research is needed to evaluate forage legume species which complement NWSG in mixture rather than compete with them. / Master of Science / Interseeding legumes in native warm-season grasses (NWSG) may improve the nutritive value of the stand, result in more consistent forage availability throughout the growing season, and increase forage yield. These benefits are often not realized due to difficulties in establishing legumes in existing NWSG stands. The objective of this study was to investigate the effect of planting method of legume interseeding, timing of legume interseeding, and the efficacy of burning residue on legume establishment in NWSG. Two forage legumes, 'Alice' white clover (Trifolium repens L.) and 'Freedom HR' red clover (Trifolium pratense L.), were interseeded into mixed 'Niagara' big bluestem (Andropogon gerardii Vitman), 'GA Ecotype' Indiangrass (Sorghastrum nutans Nash), and 'Camper' little bluestem (Schizachyrium scoparium) pasture in 2022 and 2023 at the Southern Piedmont AREC in Blackstone, Virginia. Planting method at three levels (no-till drill, broadcast, and non-planted control) were evaluated at three planting timing levels (fall planting, winter planting, and winter planting with burned residue). Among the treatment combinations, burned and drilled plots produced more clover plants in spring at 236 plants m-2, followed by winter drill (146 plants m-2) and burn broadcast (133 plants m-2). All fall plantings and all control plots were similar with a mean of 21 plants m-2. As a result of greater initial clover emergence, plots that were burned or seeded in the winter had greater clover content throughout the experiment; burned and drilled plots had over 90% clover ground cover throughout the second year. Domination of plots by clover in the second year caused yields and the proportion of NWSG in the stand to decline, with burned plots yielding 5,757 kg ha-1 compared to a winter-fall mean of 7,429 kg ha-1. Plots with greater clover content were able to sustain higher crude protein content and lower neutral detergent fiber content in both the establishment year and the second year. Though interseeding legumes benefitted nutritive values, these results suggest that red clover may be incompatible with the NWSG which were evaluated. Burned plots were especially affected by excessive competition. Further research is needed to evaluate forage legume species which complement NWSG in mixture rather than compete with them.

Lowland switchgrass (Panicum virgatum), big bluestem (Andropogon gerardii), indiangrass (Sorghastrum nutans), upland switchgrass (Panicum virgatum), little bluestem (Schizachyrium scoparium), beaked panicum (Panicum capillare), and purpletop (Tridens flavus) all show strong signs of seed dormancy which contributes to extremely poor field establishment. The objective of this work was to reduce seed dormancy by selecting individuals that exhibited reduced pre-stratification dormancy in laboratory tests. The classical breeding method of phenotypic recurrent selection was used to enhance germination. Of the three tall-stature species, lowland switchgrass made the greatest improvement in pre-stratification germination, followed by indiangrass and big bluestem. The four short stature species have shown various results after one cycle of selection at Starkville. A field emergence trial was also conducted to evaluate three cycles of breeding seed with five commercially available cultivars in which Cycle 3 seed produced more plants per hectare than any of the other cultivars or germplasm.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.