CSE 151 Intro to AI: A Statistical Approach

Syllabus

I. Introduction / Prediction problems
        A taxonomy of prediction tasks
        Linear alebra review
        Probability review

II. Generative models for classification
        The generative approach to classification
        Gaussian generative models

III. Linear prediction
        Linear regression
        Logistic regression
        Perceptron and support vector machines
        Kernels
        Multiclass classification and structured output prediction

IV. Non-linear models / Deep learning
        Nearest neighbor
        Decision trees
        Feedforward networks
        Convolutional networks

V. Representation learning
        Clustering
        Linear projections: PCA and SVD
        Embeddings and manifold learning
        Autoencoders

Discussion sections

Wed 8pm-8:50pm in Center Hall 113

Prerequisites

1. Ability to write simple programs in Python: functions, control structures, string handling, arrays and dictionaries.

2. Familiarity with basic probability, at the level of CSE 21 or CSE 103.

3. Familiarity with basic linear algebra, at the level of Math 18 or Math 20F.

Course materials

1. Programming exercises should be done in Python. I recommend using Jupyter notebooks. You will be learning to use Numpy and PyTorch.

2. There is no required text for the course. But here are some useful references. The first is available as an e-book through the library website; the rest are on reserve at Geisel:
Trevor Hastie, Robert Tibshirani, and Jerome Friedman, The elements of statistical learning (2nd edition).
Gilbert Strang. Linear algebra and its applications .
Kevin Murphy, Machine learning: a probabilistic perspective.
Richard Duda, Peter Hart, and David Stork, Pattern classification (2nd edition).

Homeworks and evaluations

There will be regular homeworks, to be turned in (typed and in PDF format) on Gradescope. These will be a mix of mathematical exercises and programming projects.

No late homeworks will be accepted; however, the lowest homework score will be dropped.

We understand that reasoning about homework problems with a group of teammates can be helpful. We fully support this. However, we require that (1) you write all your own code for problems that require coding, (2) you produce your own homework write-up, and (3) that you list the names of your teammates in your submission.

There will be five in-class quizzes. There will be no makeup quizzes; however, the lowest quiz score will be dropped.

Grading

Homeworks: 50% (lowest score will be dropped)
Quizzes: 50% (lowest score will be dropped)

Homeworks and quizzes will have point values. Assignments with more points will count proportionally more towards your total grade. For example, your final homework grade for the course will be determined by the number of homework points you recieved, divided by the total number of possible homework points. Your final grade in the course will be determined by the average of your overall homework point percentage and your overal quiz point percentage.

We will use standard grade cutoffs: A+ at instructor discretion, A 93-100%, A- 90-92%, B+ 87-89%, B 83-86%, B- 80-82%, C+ 77-79%, C 73-76%, C- 70-72%, D+ 67-69%, D 63-66%, D- 60-62%, F 0-59%. However, we reserve the right to curve grades upwards -- i.e. your final grade based on your total percentage in the course will never be a lower letter grade than determined by the cutoffs above, but may be higher depending on the grade distribution.