📚 Syllabus & Course Curriculum
AI & Machine Learning in Healthcare
Module-by-module breakdown of Prediction of Immunogenic Response using Orange: A Machine Learning Tool, from foundations to a certified capstone project.
Orange
• Widgets, channels and building a workflow without writing code
• Loading data, assigning target and meta attributes correctly
• Where a visual tool is genuinely faster and where it becomes a limitation
Data
• Peptide and antigen features, encodings and derived descriptors
• Missing values, duplicates and near-duplicate sequences across splits
• Class imbalance, since immunogenic examples are typically the minority
Exploration
• Distributions, scatter plots and correlation before modelling
• PCA, t-SNE and MDS for structure, and the over-reading they invite
• Outlier detection and deciding whether a point is error or biology
Modelling
• Logistic regression, random forest, SVM and naive Bayes side by side
• Test and Score with cross-validation, and stratification that must be set
• ROC, precision-recall and confusion matrices read correctly
Interpretation
• Feature importance, and treating it as a hypothesis rather than a mechanism
• Sequence similarity leakage between training and test as the standard trap
• Exporting a workflow reproducibly and knowing when to move to code
Earn government-registered certification in Prediction of Immunogenic Response using Orange: A Machine Learning Tool
e-Certificate and e-Marksheet issued on successful completion.
