9/3/2024 (Tuesday) Turn in Bayes homework. Quick library orientation. New notes on Linear Regression – new notes provided.
HW: Complete the linear regression notebook, (html version) for classwork/homework. As an application, do a linear regression on the London weather dataset (this part is not HW yet but will be). For next class: pick some new ideas from your Money List to discuss.
9/5/2024 (Thursday) Turn in Linear Regression homework. Go over Bayes and Regression notebooks in class. Bayes Key Discuss goodness of fit measures for categorical and quantitative data. Notes on Pearson’s Correlation Coefficient.
Then find your own dataset somewhere and perform a linear regression on it. In both cases use the LR algorithm you coded already; do not use built-in linear regression tools, please. Make your data analysis into a nice, brief write-up (emphasis on brief) and turn it in before class ends (as a jupyter notebook).
Classwork: Discuss some recent money list ideas
HW: Finish classwork, any other old notebooks that aren’t done yet
9/9/2024 (Monday) Unit Topic: Matrices in Python and Numpy. Today we will code up standard algorithms for performing matrix arithmetic operations using Python lists. Please download and modify this matrix notebook (pdf) and ftp it to the pi when you’re finished. The next notebook on Gaussian elimination will be provided later.
9/11/2024 (Wednesday) Gaussian elimination, on paper and using numpy.
Gaussian Elimination notebook (pdf). You’ll need this picture to display the Gaussian notebook properly, in the same folder. (only turn in this notebook when you finish Friday)
9/13/2024 (Friday) Complete regression analysis of error terms in Gaussian Elimination lab. Notes of log-log regression (notebook)
Note: the value of the extension is always its uniqueness. If everybody in class turns in essentially the same solution for an optional assignment, its value degrades to zero. These are a chance to explore something interesting and to be original and creative!
9/17/2024. Fire. A literal fire.
9/19/2024 (Thursday) Still closed for fire. But there is homework
Please read an article about recent results in CS, and come to class ready to talk about it. More information is here
Finish all old labs and extensions and turn them in next class. I don’t want to spend class time working on them anymore. Please email or send a Remind if you have any questions!
I’ll probably go over some/most/all labs and extensions next class.
Class activity: Work on Running Time Analysis lab, (html) in which running times of 4 different matrix multiply algorithms are considered. Discussion of pandas topics such as filtering and manipulating data and plotting using pandas.plot.
No new HW
If you are interested in the C++ matrix code check it out, along with an example of using pybind11 to call it from python.
9/25/2024 (Wednesday) Finish Running Time Analysis lab, (html). Talk about confidence intervals on regression coefficients. Introduce new lab on data ingesting and visualization with the HOBO data lab (html)
9/27/2024 (Friday) Turn in Running Time Analysis (I think everybody finished?) Work in class on your part (1,2 or 3) of the HOBO data lab. I do want to go over Running Time Analysis results and confidence intervals (new topic).
Classwork: HOBO lab, hopefully finish your part
Homework: Finish HOBO if needed, prepare some Money List ideas for next class
Next class will be short for PSAT. I have a short group activity planned for our brief time together so if you know you won’t be here, please let us know in advance if possible. Thanks!
10/1/2024 (Tuesday) PSAT day followed by a short class. We will get into groups and select articles for our next informal report. Details here.
classwork: find a dataset and model it using ‘mushroom’ as a guide
homework: finish what you started in class if you’re not happy with it!
homework: we will do book club (article club) next class. be ready to discuss in groups
HOBO: submit what you have on HOBO – I have some ideas to make this into a bigger project but we’ll put it on the back-burner for a bit and bring it back soon. Each of the 3 groups has made great progress!
10/11/2023 (Fri) More linear regression – feature selection, regularization, normalization. We learn about ridge regression and lasso regression which are enhancements to regular linear regression.
