Thashmee Karunaratne

Pre-Processing Structured Data for Standard Machine Learning Algorithms by Supervised Graph Propositionalization - A Case Study with Medicinal Chemistry Datasets

Graph propositionalization methods can be used to transform structured and relational data into fixed-length feature vectors, enabling standard machine learning algorithms to be used for generating predictive models. It is however not clear how well different propositionalization methods work in conjunction with different standard machine learning algorithms. Three different graph propositionalization methods are investigated in conjunction with three standard learning algorithms: random forests, support vector machines and nearest neighbor classifiers. An experiment on 21 datasets from the domain of medicinal chemistry shows that the choice of propositionalization method may have a significant impact on the resulting accuracy. The empirical investigation further shows that for datasets from this domain, the use of the maximal frequent item set approach for propositionalization results in the most accurate classifiers, significantly outperforming the two other graph propositionalization methods considered in this study, SUBDUE and MOSS, for all three learning methods.

Graph Propositionalization for Random Forests

Graph propositionalization methods transform structured and relational data into a fixed-length feature vector format that can be used by standard machine learning methods. However, the choice of propositionalization method may have a significant impact on the performance of the resulting classifier. Six different propositionalization methods are evaluated when used in conjunction with random forests. The empirical evaluation shows that the choice of propositionalization method has a significant impact on the resulting accuracy for structured data sets. The results furthermore show that the maximum frequent itemset approach and a combination of this approach and maximal common substructures turn out to be the most successful propositionalization methods for structured data, each significantly outperforming the four other considered methods.

Comparative analysis of the use of chemoinformatics-based and substructure-based descriptors for quantitative structure-activity relationship QSAR modeling

Quantitative structure-activity relationship QSAR models have gained popularity in the pharmaceutical industry due to their potential to substantially decrease drug development costs by reducing expensive laboratory and clinical tests. QSAR modeling consists of two fundamental steps, namely, descriptor discovery and model building. Descriptor discovery methods are either based on chemical domain knowledge or purely data-driven. The former, chemoinformatics-based, and the latter, substructures-based, methods for QSAR modeling, have been developed quite independently. As a consequence, evaluations involving both types of descriptor discovery method are rarely seen. In this study, a comparative analysis of chemoinformatics-based and substructure-based approaches is presented. Two chemoinformatics-based approaches; ECFI and SELMA, are compared to five approaches for substructure discovery; CP, graphSig, MFI, MoFa and SUBDUE, using 18 QSAR datasets. The empirical investigation shows that one of the chemo-informatics-based approaches, ECFI, results in significantly more accurate models compared to all other methods, when used on their own. Results from combining descriptor sets are also presented, showing that the addition of ECFI descriptors to any other descriptor set leads to improved predictive performance for that set, while the use of ECFI descriptors in many cases also can be improved by adding descriptors generated by the other methods.

Can Frequent Itemset Mining Be Efficiently and Effectively Used for Learning from Graph Data

Standard graph learning approaches are often challenged by the computational cost involved when learning from very large sets of graph data. One approach to overcome this problem is to transform the graphs into less complex structures that can be more efficiently handled. One obvious potential drawback of this approach is that it may degrade predictive performance due to loss of information caused by the transformations. An investigation of the tradeoff between efficiency and effectiveness of graph learning methods is presented, in which state-of-the-art graph mining approaches are compared to representing graphs by itemsets, using frequent itemset mining to discover features to use in prediction models. An empirical evaluation on 18 medicinal chemistry datasets is presented, showing that employing frequent itemset mining results in significant speedups, without sacrificing predictive performance for both classification and regression.

The effect of multiple change processes on quality and completion rate of theses: a longitudinal study

Abstract Improving the quality of Bachelor’s and Master’s theses while at the same time increasing the number of theses without expanding the existing resources proportionately is a huge challenge faced by higher educational institutions. The aim of this study is to investigate the effect of multiple change processes on Bachelors and Masters level thesis work in a selected higher educational institution. The following research questions were studied: (1) How has the thesis quality changed? (2) How has the number of completed theses changed? and, (3) How has the ratio of completed theses per supervisor changed? The change processes were introduced into the thesis process in the Department of Computer and Systems Sciences (DSV), Stockholm University during 2008–2014. The results show that the quality and the number of completed theses have significantly increased. The multiple change processes including a purpose built ICT system named SciPro, which was introduced and improved incrementally during 2010–2014 are discussed and evaluated in relation to these results.

ICT Capacity Building: A Critical Discourse Analysis of Rwandan Policies from Higher Education Perspective

Abstract With the development of technology in the 21st Century, education systems attempt to integrate technology-based tools to improve experiences in pedagogy and administration. It is becoming increasingly prominent to build human and ICT infrastructure capacities at universities from policy to implementation level. Using a critical discourse analysis, this study investigates the articulation of ICT capacity building strategies from both national and institutional ICT policies in Rwanda, focusing on the higher education. Eleven policy documents were collected and deeply analyzed to understand which claims of ICT capacity building are made. The analysis shows that strategies for building ICT capacities are evidently observed from national level policies and only in two institutional policies (KIST and NUR). Among 25 components of ICT capacity building used, the ones related to human capacity are not plainly described. Additionally, neither national nor institutional policy documents include the creation of financial schemes for students to acquire ICT tools whilst learners are key stakeholders. Although there is some translation of ICT capacity building strategies from national to some institutional policies, planning for motivation and provision of incentives to innovators is not stated in any of the institutional policies and this is a key to effective technology integration.

The Effect of Background Knowledge in Graph-Based Learning in the Chemoinformatics Domain

Typical machine learning systems often use a set of previous experiences (examples) to learn concepts, patterns, or relations hidden within the data [1]. Current machine learning approaches are cha ...