A systematic methodology to evaluating optimised machine learning based network intrusion detection systems
- Authors: Chindove, Hatitye Ethridge
- Date: 2022-10-14
- Subjects: Intrusion detection systems (Computer security) , Machine learning , Computer networks Security measures , Principal components analysis
- Language: English
- Type: Academic theses , Master's theses , text
- Identifier: http://hdl.handle.net/10962/362774 , vital:65361
- Description: A network intrusion detection system (NIDS) is essential for mitigating computer network attacks in various scenarios. However, the increasing complexity of computer networks and attacks makes classifying unseen or novel network traffic challenging. Supervised machine learning techniques (ML) used in a NIDS can be affected by different scenarios. Thus, dataset recency, size, and applicability are essential factors when selecting and tuning a machine learning classifier. This thesis explores developing and optimising several supervised ML algorithms with relatively new datasets constructed to depict real-world scenarios. The methodology includes empirical analyses of systematic ML-based NIDS for a near real-world network system to improve intrusion detection. The thesis is experimental heavy for model assessment. Data preparation methods are explored, followed by feature engineering techniques. The model evaluation process involves three experiments testing against a validation, un-trained, and retrained set. They compare several traditional machine learning and deep learning classifiers to identify the best NIDS model. Results show that the focus on feature scaling, feature selection methods and ML algo- rithm hyper-parameter tuning per model is an essential optimisation component. Distance based ML algorithm performed much better with quantile transformation whilst the tree based algorithms performed better without scaling. Permutation importance performs as a feature selection method compared to feature extraction using Principal Component Analysis (PCA) when applied against all ML algorithms explored. Random forests, Sup- port Vector Machines and recurrent neural networks consistently achieved the best results with high macro f1-score results of 90% 81% and 73% for the CICIDS 2017 dataset; and 72% 68% and 73% against the CICIDS 2018 dataset. , Thesis (MSc) -- Faculty of Science, Computer Science, 2022
- Full Text:
- Date Issued: 2022-10-14
- Authors: Chindove, Hatitye Ethridge
- Date: 2022-10-14
- Subjects: Intrusion detection systems (Computer security) , Machine learning , Computer networks Security measures , Principal components analysis
- Language: English
- Type: Academic theses , Master's theses , text
- Identifier: http://hdl.handle.net/10962/362774 , vital:65361
- Description: A network intrusion detection system (NIDS) is essential for mitigating computer network attacks in various scenarios. However, the increasing complexity of computer networks and attacks makes classifying unseen or novel network traffic challenging. Supervised machine learning techniques (ML) used in a NIDS can be affected by different scenarios. Thus, dataset recency, size, and applicability are essential factors when selecting and tuning a machine learning classifier. This thesis explores developing and optimising several supervised ML algorithms with relatively new datasets constructed to depict real-world scenarios. The methodology includes empirical analyses of systematic ML-based NIDS for a near real-world network system to improve intrusion detection. The thesis is experimental heavy for model assessment. Data preparation methods are explored, followed by feature engineering techniques. The model evaluation process involves three experiments testing against a validation, un-trained, and retrained set. They compare several traditional machine learning and deep learning classifiers to identify the best NIDS model. Results show that the focus on feature scaling, feature selection methods and ML algo- rithm hyper-parameter tuning per model is an essential optimisation component. Distance based ML algorithm performed much better with quantile transformation whilst the tree based algorithms performed better without scaling. Permutation importance performs as a feature selection method compared to feature extraction using Principal Component Analysis (PCA) when applied against all ML algorithms explored. Random forests, Sup- port Vector Machines and recurrent neural networks consistently achieved the best results with high macro f1-score results of 90% 81% and 73% for the CICIDS 2017 dataset; and 72% 68% and 73% against the CICIDS 2018 dataset. , Thesis (MSc) -- Faculty of Science, Computer Science, 2022
- Full Text:
- Date Issued: 2022-10-14
Rural households’ perceptions of an invasive alien species rosa rubiginosa l. (rosehip) and the role it plays in rural livelihoods in Lesotho
- Authors: Makhorole, Thato Violet
- Date: 2022-10-14
- Subjects: Rose hips Lesotho , Invasive plants Lesotho , Rural poor Lesotho , Probit model , Principal components analysis
- Language: English
- Type: Academic theses , Master's theses , text
- Identifier: http://hdl.handle.net/10962/403069 , vital:69919
- Description: Despite the vast research on the negative impacts of invasive alien species on the environment, these species remain part of the rural communities due to their numerous livelihood uses. Thus, more research is required, focusing mainly on the impacts of invasive alien species on the livelihoods of rural communities. This study investigated the community perceptions of rosehip (Rosa rubiginosa) and its contribution to rural communities as an invasive alien species. Four community councils, Pitseng, Matlameng, Limamarela and Mphorosane in the Leribe District Lesotho, were assessed. The study followed the pragmatism paradigm. The contribution of rosehip to rural livelihoods was analysed by comparing income from rosehip with other income sources. The study used simple random sampling and snowball sampling to select a representative of 160 respondents. The primary data was collected using semi-structured questionnaires. Moreover, SPSS and Stata statistical package programs were used for statistical analyses. The results showed that rosehip's livelihood benefits, its negative impacts, the length of time it has been available in the area, and its abundance highly influence the social, economic and environmental perception of rural communities. Furthermore, the study revealed that although the income from rosehip is extremely low and available for only three months of the year, the income plays an important part to the poorer households who have no other income sources. The study found that the main reason for engagement in rosehip harvesting despite its challenging nature was unemployment. The study also revealed that rosehip is part of the risk-reducing strategy or income diversification. Some households used it to complement other sources of income, such as agricultural production. Rosehip trade, if well-controlled, has the potential to alleviate rural poverty by creating job opportunities, providing a source of household income, and acting as a safety net in the face of shocks such as limited job opportunities and food shortages. In conclusion, households’ perceptions of rosehip have proven that rosehip is a valuable resource that provides a supplementary income that contributes towards alleviating poverty in Lesotho’s rural communities. The study recommends the private sector to establish and manages small agro-processing industries focusing on products used daily. The study also encourages environmental education and indigenous knowledge among community members, which would include knowledge and recognition of invasive alien species and their potential benefits and threats. , Thesis (MEcon) -- Faculty of Commerce, Economics and Economic History, 2022
- Full Text:
- Date Issued: 2022-10-14
- Authors: Makhorole, Thato Violet
- Date: 2022-10-14
- Subjects: Rose hips Lesotho , Invasive plants Lesotho , Rural poor Lesotho , Probit model , Principal components analysis
- Language: English
- Type: Academic theses , Master's theses , text
- Identifier: http://hdl.handle.net/10962/403069 , vital:69919
- Description: Despite the vast research on the negative impacts of invasive alien species on the environment, these species remain part of the rural communities due to their numerous livelihood uses. Thus, more research is required, focusing mainly on the impacts of invasive alien species on the livelihoods of rural communities. This study investigated the community perceptions of rosehip (Rosa rubiginosa) and its contribution to rural communities as an invasive alien species. Four community councils, Pitseng, Matlameng, Limamarela and Mphorosane in the Leribe District Lesotho, were assessed. The study followed the pragmatism paradigm. The contribution of rosehip to rural livelihoods was analysed by comparing income from rosehip with other income sources. The study used simple random sampling and snowball sampling to select a representative of 160 respondents. The primary data was collected using semi-structured questionnaires. Moreover, SPSS and Stata statistical package programs were used for statistical analyses. The results showed that rosehip's livelihood benefits, its negative impacts, the length of time it has been available in the area, and its abundance highly influence the social, economic and environmental perception of rural communities. Furthermore, the study revealed that although the income from rosehip is extremely low and available for only three months of the year, the income plays an important part to the poorer households who have no other income sources. The study found that the main reason for engagement in rosehip harvesting despite its challenging nature was unemployment. The study also revealed that rosehip is part of the risk-reducing strategy or income diversification. Some households used it to complement other sources of income, such as agricultural production. Rosehip trade, if well-controlled, has the potential to alleviate rural poverty by creating job opportunities, providing a source of household income, and acting as a safety net in the face of shocks such as limited job opportunities and food shortages. In conclusion, households’ perceptions of rosehip have proven that rosehip is a valuable resource that provides a supplementary income that contributes towards alleviating poverty in Lesotho’s rural communities. The study recommends the private sector to establish and manages small agro-processing industries focusing on products used daily. The study also encourages environmental education and indigenous knowledge among community members, which would include knowledge and recognition of invasive alien species and their potential benefits and threats. , Thesis (MEcon) -- Faculty of Commerce, Economics and Economic History, 2022
- Full Text:
- Date Issued: 2022-10-14
Targeting allosteric sites of Escherichia coli heat shock protein 70 for antibiotic development
- Authors: Okeke, Chiamaka Jessica
- Date: 2019
- Subjects: Heat shock proteins , Escherichia coli , Allosteric proteins , Antibiotics , Molecular chaperones , Ligands (Biochemistry) , Molecular dynamics , Principal components analysis , South African Natural Compounds Database
- Language: English
- Type: text , Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10962/115998 , vital:34287
- Description: Hsp70s are members of the heat shock proteins family with a molecular weight of 70-kDa and are the most abundant group in bacterial and eukaryotic systems, hence the most extensively studied ones. These proteins are molecular chaperones that play a significant role in protein homeostasis by facilitating appropriate folding of proteins, preventing proteins from aggregating and misfolding. They are also involved in translocation of proteins into subcellular compartments and protection of cells against stress. Stress caused by environmental or biological factors affects the functionality of the cell. In response to these stressful conditions, up-regulation of Hsp70s ensures that the cells are protected by balancing out unfolded proteins giving them ample time to repair denatured proteins. Hsp70s is connected to numerous illnesses such as autoimmune and neurodegenerative diseases, bacterial infection, cancer, malaria, and obesity. The multi-functional nature of Hsp70s predisposes them as promising therapeutic targets. Hsp70s play vital roles in various cell developments, and survival pathways, therefore targeting this protein will provide a new avenue towards the discovery of active therapeutic agents for the treatment of a wide range of diseases. Allosteric sites of these proteins in its multi-conformational states have not been explored for inhibitory properties hence the aim of this study. This study aims at identifying allosteric sites that inhibit the ATPase and substrate binding activities using computational approaches. Using E. coli as a model organism, molecular docking for high throughput virtual screening was carried out using 623 compounds from the South African Natural Compounds Database (SANCDB; https://sancdb.rubi.ru.ac.za/) against identified allosteric sites. Ligands with the highest binding affinity (good binders) interacting with critical allosteric residues that are druggable were identified. Molecular dynamics (MD) simulation was also performed on the identified hits to assess for protein-inhibitor complex stability. Finally, principal component analysis (PCA) was performed to understand the structural dynamics of the ligand-free and ligand-bound structures during MD simulation.
- Full Text:
- Date Issued: 2019
- Authors: Okeke, Chiamaka Jessica
- Date: 2019
- Subjects: Heat shock proteins , Escherichia coli , Allosteric proteins , Antibiotics , Molecular chaperones , Ligands (Biochemistry) , Molecular dynamics , Principal components analysis , South African Natural Compounds Database
- Language: English
- Type: text , Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10962/115998 , vital:34287
- Description: Hsp70s are members of the heat shock proteins family with a molecular weight of 70-kDa and are the most abundant group in bacterial and eukaryotic systems, hence the most extensively studied ones. These proteins are molecular chaperones that play a significant role in protein homeostasis by facilitating appropriate folding of proteins, preventing proteins from aggregating and misfolding. They are also involved in translocation of proteins into subcellular compartments and protection of cells against stress. Stress caused by environmental or biological factors affects the functionality of the cell. In response to these stressful conditions, up-regulation of Hsp70s ensures that the cells are protected by balancing out unfolded proteins giving them ample time to repair denatured proteins. Hsp70s is connected to numerous illnesses such as autoimmune and neurodegenerative diseases, bacterial infection, cancer, malaria, and obesity. The multi-functional nature of Hsp70s predisposes them as promising therapeutic targets. Hsp70s play vital roles in various cell developments, and survival pathways, therefore targeting this protein will provide a new avenue towards the discovery of active therapeutic agents for the treatment of a wide range of diseases. Allosteric sites of these proteins in its multi-conformational states have not been explored for inhibitory properties hence the aim of this study. This study aims at identifying allosteric sites that inhibit the ATPase and substrate binding activities using computational approaches. Using E. coli as a model organism, molecular docking for high throughput virtual screening was carried out using 623 compounds from the South African Natural Compounds Database (SANCDB; https://sancdb.rubi.ru.ac.za/) against identified allosteric sites. Ligands with the highest binding affinity (good binders) interacting with critical allosteric residues that are druggable were identified. Molecular dynamics (MD) simulation was also performed on the identified hits to assess for protein-inhibitor complex stability. Finally, principal component analysis (PCA) was performed to understand the structural dynamics of the ligand-free and ligand-bound structures during MD simulation.
- Full Text:
- Date Issued: 2019
- «
- ‹
- 1
- ›
- »