Cluster 12: Machine Learning: Can We Teach a Computer to Think?
Week 1
Monday (7/10)
After a tasty breakfast, students settled in for their first day of COSMOS packed with introductions, lectures, and lab time. Cluster 12 started their morning in Jacob Hall and was able to see the Geisel Library and the Falling Star sights on campus on their walk to the first lecture. During class, students began with icebreakers and shared boring facts about themselves. Of course, before diving into Machine Learning, they began by learning the fundamentals of Python. After meeting Professor Curt, Professor Niema, and the Teacher Fellow, Shirley, they had a lecture on Python Basics such as strings, variables, booleans, loops, and conditional statements. Following lunch, students were able to meet our TA’s and get started with Python challenge problems (which proved to be quite a challenge). After a long and eventful day of learning, Cluster 12 picked a skit idea for the COSMOlympics (which is a surprise!).
Tuesday (7/11)
Cluster 12 stated this day with something new–a science communication lecture on Ethics with Philosophy Professor Brandt. After learning the topic of science ethics and dilemmas posed by new innovations in science, students were introduced to the Ethic Essay everyone will be writing. Right after, Professor Niema reviewed Python basics and gave them an introduction to Machine Learning using a Cookie Scenario. During lab time students were able to complete the Python Challenge problems from Monday and get started on the Decision Trees challenge with the help of the TA’s.
Wednesday (7/12)
Wednesday started off with a presentation from Teacher Fellow Shirley about science communication. Fun fact: 30% of communication is done through tone of voice, 60% is body language, which means that only 10% is done through words! Afterwards, students continued learning about classification, which is a type of supervised machine learning that predicts an output from an input based on the data it was trained on. Professor Niema covered all the Python code in an example that predicted if an extremely picky friend would like a cookie based on the amount of chocolate chips and the diameter. In the afternoon session, Professor Niema gave a lecture about a different type of classification called k-nearest neighbors, which predicts an output from an input based on the data points closest to it. Student’s tested their knowledge by writing their own program that would output what variety a wine is based on its characteristics.
Thursday (7/12)
After much anticipation, Cluster 12 began class by choosing a topic for their ethics essay on the topic of AI in the workplace environment, and starting their drafts. Before the lab, students received their fashionable Cluster 12 t-shirts. During the lab, Professor Niema taught k-Means Clustering, which is a type of unsupervised learning that groups data into clusters. Students implemented their knowledge into the previous wine problem to try a different method to sort the wine.
