ECE 5464 Applications of Machine Learning
Applications of Machine Learning (ML) for predictive data analytics. Probability for ML including conditional probability, the product and chain rule, and the Theorem of Total Probability. Data preparation for ML algorithms, normalization, cleaning, and imputation of missing values. Information-based learning using decision trees. Similarity-based methods, data classification and clustering. Probability-based learning, conditional probability and Bayes’ theorem, and applications. Linear and logistic regression and optimization-based learning. Performance evaluation of ML systems. Artificial Neural Networks. Real-world applications of ML and case studies. Not for CPE-MS, EE-MS, CPE-PhD, or EE-PhD credit. May be taken for CPE-MEng or EE-MEng credit.
Why take this course?
Recent advances in Machine Learning (ML) algorithms and software implementations of these algorithms are essential tools in many computer and electrical engineering application areas. This course provides graduate students an introduction to the fundamental ML concepts and will prepare these students for subsequent graduate courses in specific research areas. The course focuses on the application of ML techniques through their use in real-world case studies.
- Apply standard Machine Learning (ML) approaches in real-world scenarios using software tools for predictive data analysis.
- Prepare raw data sets for use by ML algorithms and software using appropriate techniques.
- Formulate decision-tree solutions in information-based learning applications.
- Perform data classification and clustering for ML applications using similarity metrics.
- Compute probability-based solutions for inference and prediction using Bayes’ theorem.
- Apply optimization-based learning and regression techniques to engineering applications.
- Evaluate ML approaches and systems using standard performance measures for specific case studies.