Skip to main content

Machine Learning II

(Elective Course)

Machine Learning II will extend what was covered in Machine Learning I, which is a prerequisite. We will cover a variety of more advanced supervised and unsupervised learning models with a particular application to marketing communication problems. There will be a module advanced market segmentation models covering the nuances of k-means cluster along with the Gaussian mixture and latent class models. Next, the class will cover text mining models. We will cover the basic workflow (stop words, stemming, dictionaries, bigrams), word clouds, TFIDF, chi-square, sentiment analysis, topic modeling, LSI, and incorporating text variables in predictive models. There will be a module covering algorithms for recommendation systems (RS). We will go into detail on the content-based and collaborative filtering, including user-user, item-item, hybrid, and decomposition methods. I will introduce reinforcement learning. There will also be a module on neural networks. We will go into detail on feedforward neural networks, including learning with backpropagation and stochastic gradient descent. We will discuss different network architectures and training strategies. I will then give an overview convolution networks. Assignments will done in Python and R.