learning about machine learning

1 min read

Hello, I’m Parker. I’m a graduate student in the developmental division of UConn’s Psychology program and also a fellow in UConn’s IGERT program in Language Plasticity.

Before spending a summer embarking on an autodidactic journey into machine learning, I wanted to write a bit about my motivation to take the course. Below, I’ve listed seven goals for my summer-long course of study. Many of these objectives are simple and straightforward, but others are aspirational and ultimately represent larger ideas about science and developmental psychology. I quickly jotted down some ideas at the end of this post, but I may return later to formalize them once I’ve spent more time with them.

Objectives for Learning about Machine Learning

  • Obtain an understanding of the R language, its syntax, popular R packages, and IDEs
  • Understand the theoretical implications of machine and statistical learning, including their areas of overlap and uniqueness
  • Learn how to implement machine learning algorithms and models programmatically using the R language
  • Discover ways that machine learning can interface with personal research interests
  • Form collaborative relationships with fellow graduate students to engender future interdisciplinary research projects and publications
  • Acquire general technical skills that can be applied broadly across multiple research domains
  • And, finally, push the envelope and revolutionize the field of developmental psychology by embracing, borrowing, and adapting mathematical tools and techniques from non-psychology STEM-related fields

Summer Schedule

Material Due by…

Statistical Learning

5 June
Linear Regression 12 June
Classification 26 June
Resampling Methods 3 July
Linear Model Selection and Regularization 10 July
Moving Beyond Linearity 17 July
Tree-Based Models 24 July
Support Vector Machines 31 July
Unsupervised Learning 7 August

I believe that innovation in the psychological sciences, particularly developmental psychology, demands a set of tools that extends beyond what traditional training programs typically offer. To truly innovate and advance scientific theory, today’s developmental psychology students need to emerge from their graduate studies equipped with a broad range of analytical techniques and multi-disciplinary research experience(s). On these grounds, It seems that many graduate-level psychology programs favor knowledge acquisition over skills acquisition. Unfortunately, this is often to students’ detriment. I hope to remediate this by obtaining a general set of technical skills for modeling, analysing, and visualizing data on my own. I begin, first, with machine learning.

a year of quiet

Henry Wolf
1 min read

individual differences in sensitivity to homophony in visual word…

I successfully defended my MS thesis, titled Individual Differences in Sensitivity to Homophony in Visual Word Recognition, on 24 April 2017. Follow the link...
Henry Wolf
20 sec read

a summer of modeling and mastery

I spent last summer in Suzhou studying Mandarin. 可是,我忘了多中文。でも、実は今は日本語勉強したい。I will spend this summer in Connecticut doing research and writing. While it is technically the...
Henry Wolf
1 min read

2 Replies to “learning about machine learning”

  1. I hear you on the knowledge over skills bias. There is a large expectation that one will learn the skills in their lab, leaving students with a limited skill set. I am glad that we are breaking out of that mold.

    Here’s to collaborative relationships!

    1. Agreed. Even the knowledge that was once reserved to the halls and libraries of universities is now freely available on Coursera, Kahn Academy, and similar sites. Anyone with an internet connection can take college-level classes in history, linguistics, mathematics, etc. I think this will radically change the academic value of knowledge and skills in the future, tipping the scales towards the latter: academics will become doers instead of just knowers.

Leave a Reply

Your email address will not be published. Required fields are marked *