Faculty of Science (ETDs)

Permanent URI for this community

https://hdl.handle.net/10539/37933

Browse

Now showing 1 - 10 of 10

Analyzing the performance and generalisability of incorporating SimCLR into Proximal Policy Optimization in procedurally generated environments
(University of the Witwatersrand, Johannesburg, 2024) Gilbert, Nikhil; Rosman, Benjamin
Multiple approaches to state representation learning have been shown to improve the performance of reinforcement learning agents substantially. When used in reinforcement learning, a known challenge in state representation learning is enabling an agent to represent environment states with similar characteristics in a manner that would allow said agent to comprehend it as such. We propose a novel algorithm that combines contrastive learning with reinforcement learning so that agents learn to group states by common physical characteristics and action preferences during training. We subsequently generalise these learnings to previously encountered environment obstacles. To enable a reinforcement learning agent to use contrastive learning within its environment interaction loop, we propose a state representation learning model that employs contrastive learning to group states using observations coupled with the action the agent chose within its current state. Our approach uses a combination of two algorithms that we augment to demonstrate the effectiveness of combining contrastive learning with reinforcement learning. The state representation model for contrastive learning is a Simple Framework for Contrastive Learning of Visual Representations (SimCLR) by Chen et al. [2020], which we amend to include action values from the chosen reinforcement learning environment. The policy gradient algorithm (PPO) is our chosen reinforcement learning approach for policy learning, which we combine with SimCLR to form our novel algorithm, Action Contrastive Policy Optimization (ACPO). When combining these augmented algorithms for contrastive reinforcement learning, our results show significant improvement in training performance and generalisation to unseen environment obstacles of similar structure (physical layout of interactive objects) and mechanics (the rules of physics and transition probabilities).
Clustering and Classification Techniques in the Presence of Outliers: An Application to the Johannesburg Stock Exchange Stocks
(University of the Witwatersrand, Johannesburg, 2024) Maphalla, Retsebile; Chipoyera, HW
In this study, the impact of outliers on clustering using the K-means algorithm was explored. It was observed that a high prevalence of outliers can seriously compromise the results of clustering. A novel algorithm called Clustering-quality-aided outlier detection (CQAOD) is proposed in this study. The novelty stems from the fact that apart from identifying outliers, good quality clustering is achieved and the “optimal” number of clusters for K-means clustering of multivariate Gaussian data is simultaneously proffered. In the case of the Johannesburg Stock Exchange (JSE) data, an investigation to compare the efficacy of the following clustering techniques: Hierarchical clustering, spectral clustering, Clustering Large Applications (Clara), Density-based spatial clustering of applications with noise (DBSCAN) was done with the aim of constructing a diversified stock portfolio. The study found that the hierarchical clustering algorithm is the best algorithm to cluster the shares on the JSE
Comparison of Saccharomyces cerevisiae and the novel wild yeast used in beer fermentation and their future in industrial biotechnology
(University of the Witwatersrand, Johannesburg, 2024) Zviuya, Patience; Moodley, Sanchia; Rumbold, Karl
The alcohol industry has grown over centuries due to the increase in alcohol demand by customers. Beer consumers now understand the brewing process and they well understand the role of yeast in fermentation. The alcohol industry now has a lot of customers because of the growth that this sector has made, and this has resulted in an increased demand for new beer styles and high flavor profile beers. Due to this demand, research has been conducted on unconventional wild yeasts that can be employed in making beer, demonstrating the variety of fermentation yeasts that are available and capable of enhancing beer quality and producing a wide range of new beer varieties. The commercial S. cerevisiae yeast and eight wild yeasts (Samson’s Saison, Ragnarok, Dark knight, B. brusc, Neipa, The Proletariat, La Trappist and B. clauss) were used to ferment pale ale in different fermentation vessels. The commercial yeast was the control of this research because this yeast has been used for generations in the brewing industry. The wild yeast strains used were identified using sanger sequencing and seven of these yeasts were S. cerevisiae wild species with similarity index of more than 80% and one was B. bruxellensis strain similarity index more than 95% before and after fermentation. The research outcomes demonstrated that most of the wildyeast performed the same as the commercial yeast in terms of physical and chemical parameters however most of the wild yeast produced more volatiles and esters as compared to the commercial brewing yeast. Commercial S. cerevisiae produced the highest alcohol content 4.5% and the average alcohol content for wild yeast was 3.4% because they are challenging to regulate during fermentation and have low alcohol tolerance unlike the commercial yeast that has been harvested and used for generations. Overall, the utilization of unconventional wild yeast to make beer was identified as a promising alternative to produce beers with exotic flavours and alcohol-free beers. Future work identifying specific yeast that suit different fermentation processes and beer types are recommended.
Evaluation of the JSE’s environmental reporting requirements of South African listed companies
(University of the Witwatersrand, Johannesburg, 2024) Hariram, Viratha; Schwaibold, Ute
Through its mandatory environmental reporting requirements, the Johannesburg Stock Exchange (JSE) plays a pivotal role in the private sector to align to Environmental, Social and Governance frameworks and disclose information of a company’s environmental priorities and performance in addressing areas of concern. While there are attempts to safeguard the environment from damage and degradation, it is unclear if this framework is suitable at appropriately addressing the environmental areas of concern facing South Africa. In order to evaluate the JSE’s environmental requirements on listed companies, this study aimed to identify the local and global environmental priorities via the South African State of Environment Outlook Report (local view), National Development Plan (local view) and Sustainable Development Goals (global view) and thereafter determine if the Global Reporting Initiative, the only set of mandatory environmental requirements of the JSE, was adequately addressing the indicators they outlined. Furthermore, using a scoring system from zero to four, the study evaluated the alignment of ten randomly selected JSE listed companies against the requirements of the Global Reporting Initiative to note their level of adherence and alignment to the South African State of Environment Outlook Report, National Development Plan and Sustainable Development Goals. The results indicate that the Global Reporting Initiative (GRI) covers 73% of environmental priorities discussed in the South African State of Environment Outlook Report, National Development Plan and Sustainable Development Goals. Of the total 74 indicators of environmental concern identified from the three reports, the GRI did not require disclosure for 20 indicators. Only three indicators that were required to be disclosed by the GRI had scored a four on the rating scale due to it being an integrated disclosure that takes into account one or more other related environmental categories. The evaluation of the sampled listed company’s adherence to the GRI via their sustainability reports and integrated annual reports indicated that none of the companies had provided sufficient disclosures to meet the requirements of the GRI. 80% of the sampled listed companies made a strong alignment to the Sustainable Development Goals in their sustainability reports and / or integrated annual reports. A shortcoming of the JSE that was stated by all four sustainability experts interviewed was the lack of enforcement for listed companies to make quality and comprehensive disclosures or accountability expected from listed companies. There is no formal process of review or consequences for listed companies who do not comply.
Exploration Geology of Structurally Controlled Sediment-Hosted Gold Mineralization at Kasenseli, Northwestern Zambia
(University of the Witwatersrand, Johannesburg, 2023-10) Mulenga, Kapalakasha; Kinnaird, Judith; Nex, Paul
Occurrences of economic quantities of gold within the Roan and Nguba groups of the Katanga sedimentary units are not a common phenomenon. This is a new area of exploration for gold, and very little exploration work has been done to act as a guide in such a geological environment. More common are post-Nguba unit or Kundelungu group related gold occurrences which respectively occur around the Kasempa and Mumbwa areas and these are related to the Hook granite as well as the major shear zones that traverse these areas. The study area lies within the Lufilian Arc Fold and Thrust Belt and is located on the western end of the external fold and thrust zone which is part of the western extension of the Zambian Copperbelt. The area covering Mwinilunga is subdivided into four terranes based on regional structural features and metamorphism. The four terranes are; the Kasai Shield, the Kabompo Dome, the Western Foreland and the Lufilian Arc. Four main units that are present within the study area include the Roan member shale unit, the arkosic sandstone and its related sub-units which occur as lenses within this main unit, the Grand Conglomerat marker unit (diamitite), and the Nguba member shale which is interbedded with siltstone in some zones. Based on the contact relationship of the units to Grand Conglomerat marker unit, the oldest unit within the study area is the Roan shale unit (which could be of Mwashya age or older), followed by the arkosic sandstone, the diamictite unit which is the Grand Conglomérat marker unit and finally the Nguba shale. Polished thin section and mineral liberation analysis was undertaken on most of these units. Among the notable features of the stratigraphy in the area is the presence of a redox boundary between the pinkish-grey coarser arkosic sandstone unit (oxidising facies) and the greenish-grey clast-poor diamictite (reducing facies) where richer pockets of gold mineralisation have been observed. Gold mineralisation has been mainly observed to be hosted in the clast-poor diamictite mostly in the alteration halo which forms around the arkosic sandstone-quartz vein-diamictite contact. Further away from this contact and away from other related structures such as faults, veins and folds, mineralisation gradually diminishes. Gold mineralisation in this area shows a strong correlation to copper with the high- grade zones having elevated copper mineralisation both in the mineralised zone and in the residual soils. On a district scale, the study area and its surroundings are extensively folded and faulted, and the area has undergone episodic deformation resulting in the development of various structures. Among the notable structures within the study which are relatable to the Lufilian Arc Fold and Thrust Belt (LAFTB) on district to regional scale include folding, shear faulting, normal faulting, and the development of joints. The structural aspect plays a critical role with regards to the location of mineralisation. Observed structural features which were critical as pathways and deposition site of uriferous/mineralised fluids include first order folding with a NNE – SW oriented axial trace as well as first order NE – SW oriented shear faults and shear zone, the second order NW – SE oriented faults and veins which are the structures that host the mineralisation, and the NW – SE oriented (axial trace) folding where saddle-reef hosted mineralisation has been observed within the study. The physiochemical redox boundary between the Nguba member diamictite (reducing facies) and pre-Nguba arkosic sandstone (oxidised facies) played a critical role in the location of the mineralisation in the area. One thousand six-hundred and fifty (1650) soil samples were collected from the entire study area which covers a surface area of about 13 square km. Analysis and interpretation of soil assays showed that only copper may reliably be used as pathfinder in the exploration for gold mineralisation and/or deposits particularly in Nguba group member units. In addition, based on stratigraphic correlation, the possibility of Kamoa-style copper mineralisation has been suggested.
Mapping and assessment of informal settlements using object-based image analysis, a case study of Mamelodi, Tshwane, South Africa
(University of the Witwatersrand, Johannesburg, 2024) Mudau, Naledzani; Mhangara, Paida
The social and environmental challenges faced by people living in informal settlements or slums are widely recognized by development agendas including United Nations Sustainable Development Goals, Agenda 2063 and National Development Plans. The study aims to investigate informal settlement dynamics and spatial characteristics to generate an understanding of housing informality and environmental conditions for designing innovative sustainable solutions. The study assessed the use of 12 spectral indices and textural measures, and object-based image analysis (OBIA) technique to detect informal settlements from WorldView 2 images. A growth indicator that uses informal settlement extent and impervious surface was developed and used to assess informal settlement growth patterns between 2005 to 2020. Unmanned aerial vehicle image products, and landscape metrics were used to assess the spatial characteristics and patterns of backyard shacks and free-standing informal settlement structures. In addition, a settlement surface ecological index was developed and used to assess the ecological conditions of informal settlements. Lastly, the assessment of the location characteristics of informal settlements was done using ancillary data. The results show that the use of built-up index, coastal blue index and first order statistics mean textural measures and OBIA technique detected informal settlements with producer and user accuracies of 95% and 82% respectively. The developed informal settlement growth assessment indicator shows that informal settlement in 2020 had a slightly lower density of impervious surfaces than in 2005. The Euclidean Nearest-Neighbour Distance, Aggregation Index and Cohesion Index show that backyard shacks are less connected, less dense, and more isolated than freestanding informal settlement structures. Some informal settlements have better surface ecological conditions than some of the formal settlements. A higher extent of informal settlements continued to develop closer to formal settlements, rivers and railway lines between 2015 and 2020. The information demonstrated in this study can be used by local authorities to better understand and manage informal settlement developments, prioritize settlement upgrade projects and improve the environmental conditions and resilience of informal settlements
Regime Based Portfolio Optimization: A Look at the South African Asset Market
(University of the Witwatersrand, Johannesburg, 2023-09) Mdluli, Nkosenhle S.; Ajoodha, Ritesh; Mulaudzi, Rudzani
Financial markets change their properties (i.e mean, volatility, correlation, and distribution) with time. However, traditional portfolio optimization strategies seek to create static, all weather portfolios oblivious to this and current economic conditions. This produces portfolios that are unable to predict events with excessive skewness and kurtosis. This research investigated the difference in portfolio percentage return, of portfolios that incorporate regimes against one that does not. HMMs, binary segmentation, and PELT algorithms were used to identify regimes in 7 macro-economic features. These regimes, with regimes identified by the SARB, were incorporated into Markowitz’s mean-variance optimization technique to optimize portfolios. The base portfolio, which did not incorporate regimes, produced the least return of 761% during the period under consideration. Portfolios using HMMs identified regimes, produced, on average, the highest returns, averaging 3211% whilst the portfolio using SARB identified regimes returned 1878% during the same period. This research, therefore, shows that incorporating regimes into portfolio optimization increases the percentage return of a portfolio. Moreover, it shows that, although HMMs, on average, produced the most profitable portfolio, portfolios using regimes based on data-driven techniques do not always out-perform portfolios using the SARB identified regimes.
The impact of fly ash on the growth rate of maize (Zea mays L.) and the potential for fly ash to be used commercially in agriculture
(University of the Witwatersrand, Johannesburg, 2022) Riba, Mpho Mahlatse; Scholes, Mary
South Africa’s coal-fired power plants (CFPPs) produce about 34.4 million tons of fly ash (FA) per annum, as a by-product of the coal combustion process, which is stockpiled adjacent to these CFPPs. This FA contains traces of heavy metals (HMs), as a result it is a potential environmental hazard. The purpose of this study was, therefore, to establish the effects of (‘FA’) on the growth of maize (Zea mays L.), as well as its potential commercial use in agriculture due to its nutrient content, instead of maintaining the status quo. Another aim of the study was also to determine the phytoextraction abilities of the HMs (Mn, Cr, Ba and Mo) from contaminated soils by the maize plants. The study was conducted by planting maize seeds in five soil substrates with 0 %, 20 %, 40 %, 55 % and 60 % FA treatment, on mass basis. The ANOVA indicated that there were significant differences when comparing most of the mean stem length, number of leaves, root length and the mean mass of total plant dry biomass, across all the FA treatments, after 122 days of growth. The mean stem lengths after 122 days of growth were 86.67  20.55 cm, 148.33  13.87 cm, 154.17  18.10 cm, 121.17  14.99 cm and 149.50  24.20 cm for maize plants grown in 0 % , 20 %, 40 %, 55 % and 60 % FA treatments, respectively. The mean number of leaves were 12  1.87, 16  2.34, 18  2.83, 12  1.87 and 14  1.79 across all treatments, respectively. The key finding was that the growth of these plants improved with the amount of FA in the soil, with the plants that were grown in the 40 % FA treatment yielding the best results. The same observation was made with the root length and the mean masses of total plant biomass (roots, stems and leaves). Another important finding was that the accumulation of the HMs (Mn and Cr) in the plants increased, across the FA treatments, while their concentrations in the soil decreased across the FA treatments. The accumulation of Ba and Mo in the plants also increased, across the FA treatments, while Ba remained slightly constant and Mo increased in the soil. This indicated that the maize plants can potentially remediate Mn and Cr contaminated soils, through phytoextration, but are retained in the root system.
The reporting of the sustainability goals and the Sustainable Development Goals (SDGs) in the context of the private sector in South Africa
(University of the Witwatersrand, Johannesburg, 2022) Mosi, Tumelo
Sustainability challenges continue to mount and the fate of this planet lies in a precarious position. In response, the United Nations has introduced a global agreement known as the Sustainable Development Goals (SDGs). These are 17 goals underpinned by 169 targets for addressing social, economic and environmental issues on a global scale. The SDGs seek to do so by unlocking the potential that lies within the private sector and more. The aim of my research was to assess the sustainability goals disclosed by the top 25 Johannesburg Stock Exchange (JSE) listed companies against that of five other countries in 2017. Furthermore, I compared the environmental SDG disclosure of five companies from the metals and mining sector as well as the consumer products sector respectively, within the same period. The sample was selected from the original sample of the top 25 JSE listed companies. Comparisons were done by analysing sustainability/annual/integrated reports of the top 25 JSE listed companies. The results for the sample showed South Africa to be lagging behind the other countries with regards to disclosing sustainability goals in the corporate sustainability reporting for 2017. From 2017 to 2019, the number of reports addressing environmental SDGs from the selected sectors showed an increase. Furthermore, the study companies from the metal and mining sector performed better than those from the consumer products sector with regards to environmental SDG disclosure. However, the environmental SDG score of the study companies from the consumer products sector increased relatively more within the three years of assessment (2017-2019). Variation was also found between the study companies from the two sectors with regards to preference for specific environmental SDGs which included Responsible Consumption and Production (SDG 12), Climate Action (SDG 13), Life Below Water (SDG 14) and Life on Land (SDG 15). Moreover, the iii interviews conducted with the relevant participants revealed that SDG awareness has improved overtime, however internalisation of the SDGs remains the biggest challenge.
Water Quality Monitoring of Biological and Chemical Pollutants into the Wakkerstroom Wetland
(University of the Witwatersrand, Johannesburg, 2024) Klinkert, Celine; Scholes, Mary
Wetlands are among the most important ecosystems in the world due to the array of ecosystem services that they provide, such as flood attenuation, ground water restoration, water filtration and pollutant removal, climate change regulation and reducing erosion. The Wakkerstroom wetland in the Mpumalanga Province South Africa, plays a major role in improving downstream water quality. The wetland is threatened due to excess nutrient loading and microbiological contaminants entering the water column from a sewage ground runoff and a malfunctioning wastewater treatment plant. This study aims to determine the water quality and nutrient removal from the wetland by assessing chemical parameters and coliform and E.coli contamination as faecal indicators. Eight sampling sites were chosen along the length of the wetland, and water sample collection occurred once a month between August 2021 and July 2022. The results showed a neutral pH between 6 and 8 with no spatial or temporal significant differences. The mean Total Dissolved Solids (TDS) ranged between 50 – 145 mg/L. The mean nitrate concentrations (NO3-) ranged between 1.50 and 7.00 mg/L across the sampling sites. During the sampling period, a net total of 17.53 mg/L of nitrate was removed from the water column by the wetland. The mean phosphate concentration (PO43-) ranged between 0.40 – 1.40 mg/L, showing a total net removal of 2.14 mg/L during the sampling period. Coliform abundance was high closest to the wastewater treatment plant and sewage runoff and decreased along the length of the wetland, with an estimate removal of 1.17, calculated from a qualitative colour changing scoring system. E.coli bacteria was present in the wetland at all of the sampling sites between March 2022 to July 2022. This study resulted in the confirmation of a healthy, functioning wetland system where the wetland is able to reduce excess nutrients and coliform contamination. The wetland water quality shows similarities to studies conducted on wetlands in South Africa. The wetland is still under threat through possible reduced functioning as a result of excessive exposure to untreated domestic and sewage waste. A need exists for improved management and more stringent standards of water quality control from the wastewater treatment plant to reduce the risk of wetland degradation through excessive sewage exposure.

Browse

Browsing Faculty of Science (ETDs) by SDG "SDG-8: Decent work and economic growth"

Results Per Page

Sort Options