Prediction of learning outcomes via clickstream data using machine learning

Abstract

I have reviewed the content and presentation style of this thesis and declare it is free of plagiarism and of sufficient grammatical clarity to be examined. To the best of my knowledge, the research and writing are those of the candidate except as acknowledged in the Author Attribution Statement. I confirm that the investigations were conducted in accord with the ethics policies and integrity standards of Nanyang Technological University and that the research data are presented honestly and without prejudice.

Machine Learning Reveals Serum Glycopatterns as Potential Biomarkers for the Diagnosis of Nonalcoholic Fatty Liver Disease (NAFLD)

“…Compared with RF, GBDT adopts a stepwise iteration method; each tree is built based on the residue of the previous tree, which can better fit the complex relationships in the data. In GBDT, as the number of samples increases, the deviation of the cross-validation set has always been far from the deviation of the testing set, and the deviation of the testing set is constantly rising, which means that increasing the data set will optimize our algorithm to a certain extent …”

Machine learning approach in non-intrusive monitoring of tool wear evolution in massive CFRP automatic drilling processes in the aircraft industry

“…Mohanraj et al [55] developed a tool condition monitoring system for end milling process using different machine learning algorithms such as Decision Trees, Kernel Bayes, Multilayer Perceptron, and k-Nearest Neighbors (kNN) with accuracies up to 89%. Twardowski et al [56] monitored tool edge conditions using acoustic emission signals and a machine learning approach to Adapted from [63].…”

“…1 shows how the scale drives machine learning and deep learning progress. In this case, the performance of the model will be given not by the model itself but by the ability to select the right features as inputs, tuning the parameters of the model used [63]. Also remarkable is the fact that Deep Learning algorithms such as Neural Networks are much more complex to adjust because of the number of hyperparameters available for tuning, with a noticeable computational cost.…”

Artificial intelligence in real-time diagnostics and prognostics of composite materials and its uncertainties—a review

“…It involves offline creation/learning and testing of models, along with the online real-time operation of those models, and in some cases, real-time updates of models using new data seen in the field. Readers are strongly recommended to read the cited books for a deeper understanding of AI methods [15][16][17].…”