About W Booth School

SEP 786 Artificial Intelligence and Machine Learning Fundamentals

• Solving Problems using AI: Searching, optimization, online search agents. Constraint satisfaction.
• Knowledge, Reasoning and Planning: Logic and Inference, Planning and Acting, Knowledge Representation. Knowledge and Reasoning with Uncertainty. Machine learning problems, training and testing, overfitting.
• Modelling strategies: data preprocessing, overfitting and model tuning. Measuring predictor importance. Factors that Can Affect Model Performance. Feature selection. Measuring performance of classification models.

SEP 787  Machine Learning : Classification Models

• Discriminant analysis and other linear classification: Logistic regression, Linear discriminant analysis, Partial least squares discriminant analysis, Penalized models, Nearest shrunken centroids
• Linear Support Vector Machine: Empirical vs structural risk minimization, Soft margin classifier
• Nonlinear classification models: Nonlinear discriminant analysis, Flexible discriminant analysis, Support vector machines, K-nearest neighbours, Naïve Bayes
• Classification Trees and Rule based modes: Basic classification trees, Rule-based models, Bagged trees, Random forests, Boosting
• Remedies for severe class imbalance: The effect of class imbalance, Model tuning, Alternate cutoffs, Adjusting prior probabilities, Unequal case weights.

SEP 788 Neural Networks and Development Tools

• Introduction to AI landscape.
• Tools for building a machine learning project (Python crash course).
• Neural Networks Mathematical background.
• Hyperparameters, Generalization, Training process.
• Using Python scikit-learn.
• Project 1 – build and train a neural network using python scikit-learn
• Neural networks vs deep learning.
• TensorFlow
• Applying TensorFlow for Machine learning,
• Project 2- use TensorFlow/Keras for defining and training a deep learning network (use case).
• Other DNN frameworks and hardware.
• Project 3- More complex deep learning challenge (using Google cloud or Azure) .

SEP 789 Deep Learning and Its Applications

• Image processing using machine learning.
• Convolutional Neural Network (CNN) Architecture and training.
• Image recognition application.
• Transfer learning, Image classification, Image segmentation, Image labelling, Clinical imaging.
• Project 1 – build and train a CNN network (image processing).
• Recurrent Neural Network (RNN) Architecture; RNN Applications (Time series, Stock market).
• Deep learning application in Natural language processing.
• Smart manufacturing: Comparison between deep learning and traditional machine learning, Product quality inspection, Fault assessment, Defect prognosis.
• Project 2 – build and train a RNN network.
• Reinforcement learning: Architecture; Application in Robotics, Application in display advertising.
• Advanced topics: Autoencoders, Deep Residual Networks, Generative Models (GANs).
• More Applications: Digital & Smart Systems, Energy, Transportation, Micro-Nano Systems, Advanced Manufacturing
• Project 3 – select one project out of 3-4 themes with data sets to apply deep learning with minimal guidance.

SEP 767 Multivariate statistical methods for Big Data analysis and process improvement

• Course overview and introduction
• Principal Component Analysis (PCA)
• Applications of PCA
  • Monitoring processes
  • Data visualization
• Projection to Latent Structures (PLS)
• Applications of Latent Variable Methods
  • Soft Sensors
  • Advanced pre-processing
• Working with images
• Classification
• Dealing with batch data
• Multiblock data analysis