PREDICTING OUTCOME OF GAD-ALUM TREATMENT IN TYPE 1 DIABETES USING ELASTIC-NET AND RELIEF-BASED ALGORITHMS
Christoffer Henriksson
Avancerad (30hp)
kl 10:15, Alan Turing (In English)
[Abstract]The GAD-alum treatment is used to slow down the progression of T1D by immunomodulating the antigen GAD65. In studies of GAD-alum, some patients have shown positive results. The characteristics which enable the positive reaction have however not been identified. During studies of GAD-alum, data has been gathered in form of biological markers and other non-biological information about the patients. We investigated whether we can predict the progression of the disease using machine learning. Elastic-Net and Reliefbased algorithms have been used in order to predict the biomarker C-peptide, which is highly correlated to insulin. Two experiments where performed in order to simulate how well we can predict the progression of the disease at baseline and during the study. Results indicated that models predicting from baseline measurements did not perform well. Data available at baseline, might therefore not be enough for accurate predictions. Models predicting during the study did however perform better and could potentially be used to indicate the patients response to the treatment. |
General-purpose maintenance planning using deep reinforcement learning and Monte Carlo tree search
Viktor Holmgren
Avancerad (30hp)
kl 13:15, Alan Turing (In English)
[Abstract]Maintenance planning and execution is increasingly important for the modern industrial sector. Maintenance costs can amount to a major part of industrial spending.However, it is not as simple as just reducing maintenance budgets. A balance must be struck between risking unplanned downtime and the costs of maintenance efforts, in order to keep the profit margins needed to compete in the global markets of today. One approach to improve the effectiveness of industries is to apply intelligent maintenance planners. In this thesis, a general-purpose maintenance planner based on Monte-Carlo tree search and deep reinforcement learning is presented. This planner was evaluated and compared against two different periodic planners as well as the oracle lower bound on four different maintenance scenarios. These four scenarios are all based on servicing wind turbines. All scenarios include imperfect maintenance actions, as well as uncertainty in terms of the outcomes of maintenance actions. Furthermore, the four scenarios include both single and multi-component variants. The evaluation showed that the proposed method is outperforming both periodic planners in three of the four scenarios, with the forth being inconclusive. These results indicate that the maintenance planner introduced in this paper is a viable method, at least for these types of maintenance problems. However,further research is needed on this topic of maintenance planning under uncertainty. More specifically, the viability of the proposed method on a more diverse set of maintenance problems is needed to draw any clear general conclusions. Finally, possible improvements to the training process that are discussed in this thesis should be investigated. |
Latency and Dependability Comparison b/n Cloud and Fog Computing For Location Aware Application
Simon Mehari
Avancerad (30hp)
kl 10:15, Alan Turing (In English)
[Abstract]This thesis presents an architectural design for location aware fire safety application (Fire-fly) based on cloud and fog computing. The Fire-fly application is designed to assist the residents of a building to be notified and evacuate individually in a fire emergency that is happening in the building by the help of sensor readings collected from within the building. A part of this thesis explains how this application is implemented using different types of communication models between different clients and protocols for both cloud and fog architectures.
When it comes to communication between different clients, the fog architecture might be a preferred choice because it has lower latency than the cloud based architecture. However this preference is questioned when it comes to dependability aspect. Therefore, this thesis also studies the dependability trade-off the fog architecture has by studying the availability of the fog architecture in a simulated failure of part of the fog architecture. The system restoration period of the fog architecture is measured by suddenly shutting down parts of the fog structure that are most likely expected to be exposed to physical damages because of fire.
The following results are achieved after testing the whole system on a test site: the average time it takes for data that contains location information from the resident client to arrive at the fire brigade client for the cloud infrastructure was 277 ms and for the fog was 14 ms. And the results achieved from the service restoration period will approximately take on average 530 ms for a resident client to reconnect to a another functioning fog-cell in case of disconnection due to fire, and it takes on average 940 ms to reconnect to the cloud. |
Studying the effectiveness of dynamic analysis for fingerprinting Android malware behavior
Viktor Regard
Avancerad (30hp)
kl 13:15, Donald Knuth (In English)
[Abstract]Android is the second most targeted operating system for malware authors and to counter the development of Android malware, more knowledge about their behavior is needed. There are mainly two approaches to analyze Android malware, namely static and dynamic analysis. Recently in 2017, a study and well labeled dataset, named AMD, consisting of over 24,000 malware samples was released. It is divided into 135 varieties based on similar malicious behavior, retrieved through static analysis of the file classes.dex in the APK of each malware, whereas the labeled features were determined by manual inspection of three samples in each variety. However, static analysis is known to be weak against obfuscation techniques, such as repackaging or dynamic loading, which can be exploited to avoid the analysis. In this study the second approach is utilized and all malware in the dataset are analyzed at run-time in order to monitor their dynamic behavior. However, analyzing malware at run-time has known weaknesses as well, as it can be avoided through, for instance, anti-emulator techniques. Therefore, the study aimed to explore the available sandbox environments for dynamic analysis, study the effectiveness of fingerprinting Android malware using one of the tools and investigate whether features from AMD and the dynamic analysis correlate. For instance, by an attempt to classify the samples based on similar dynamic features and calculating the Pearson Correlation Coefficient (r) for all combinations of features from AMD and the dynamic analysis. The comparison of tools for dynamic analysis, showed a need of development, as most popular tools has been released for a long time and the common factor is a lack of continuous maintenance. As a result, the choice of sandbox environment for this study ended up as Droidbox, because of aspects like ease of use/install and easily adaptable for large scale analysis. Based on the dynamic features extracted with Droidbox, it could be shown that Android malware are more similar to the varieties which they belong to. The best metric for classifying samples to varieties, out of four investigated metrics, turned out to be Cosine Similarity, which received an accuracy of 83.6% for the entire dataset. The high accuracy indicated a correlation between the dynamic features and static features which the varieties are based on. Furthermore, the Pearson Correlation Coefficient confirmed that the manually extracted features, used to describe the varieties, and the dynamic features are correlated to some extent, which could be partially confirmed by a manual inspection in the end of the study. |
Swedish web site
