Day 24

Tuesday, May 2, 2017

Recognition by multilayer networks: deep learning I

Excellent explanation by Brandon Rohrer: (the 13 minutes up to discussion of gradient computation that we can skip, then from 21:34 to 23:30).

rohrer1 slides here

Quiz: "rohrer1" As we watch and discuss the video, each student should ask at least one question via the quiz form.

Convolutional neural networks

This weekend, please carefully watch this tutorial by Brandon Rohrer, and ask a question to Quiz: "rohrer2".

rohrer2 slides here

TensorFlow II

Another great tutorial by Magnus Erik Hvass Pedersen. Download the Jupyter notebook supplied with it.

Feature extraction vs deep learning

Compare Leafsnap

Extended data-collection project

Due 11:59pm, Friday, May 12.

Analyze the data you have been collecting throughout the semester. If you have not collected your own data, you may do this:

How good or bad are our local weather forecasts?

Here is a collection of Buffalo weather forecasts for a large portion of 2015 (which I collected for a previous edition of MTH 448/563). It is 17,000 files, 285MB expanding to 2.0GB. Your task is to choose one aspect of the forecasts (high and/or low temperature is probably the easiest to deal with) compare what was predicted with what actually happened. I hope you can invent a visual representation that clearly shows the accuracy/inaccuracy of the forecasts. Forecasts go out to 10 days ahead: the quality presumably gets worse the further out you go.

Looking ahead

Thursday, Geographical Information System software QGIS. Please install before Thursday.

Tuesday, May 9: More QGIS

Thursday, May 11: Rachael Hinkle, UB Political Science Dept. Analyzing the language of judicial opinions.

Summer part-time position available

Machine learning project related to agriculture

cassava.png