6.S077 with Professor Berwick

From 6.034 Wiki

(Difference between revisions)
Jump to: navigation, search
m
Line 1: Line 1:
-
= Introduction =
+
= Prospectus =  
-
The 2017 edition of 6.034 offers an optional 3-unit add-on set of lectures given by Professor Robert C. Berwick.
+
This term we will again offer an optional 3-unit undergraduate add-on to 6.034, 6.S077, taught by Prof. Robert C. Berwick. It will focus on scientific questions in the field and biological intelligence: what makes animals, including us, smart? How do we and animals learn? There will be a special emphasis on language.
-
This 3-unit section, 6.S077, supplements the main lectures and recitations, focusing on the ''science'' side of 6.034. In particular, 6.S063 addresses long-standing scientific questions and biological intelligence, rather than existing tools for building applications. The topics of '''evolution''' and '''human language''' are heavily emphasized.
+
Lectures will be Wednesdays 3-4 pm in room 36-153 and then repeated Fridays 2-3 pm in room 36-155; you can go to whichever lecture works for you. Additional machinery for signing up will be posted soon on the main course Wiki. There will be a lecture the first week of class on both Wednesday and Friday to introduce the add-on so that you can decide whether or not you want to pursue it.
-
= Logistics =
+
This year, one overarching theme will be how animals, including us, use “Small Data” rather than “Big Data” for learning, along with built-in constraints evolution has provided.
-
We meet every Friday, at both 2-3pm and 3-4pm (you may choose which section to attend), in 56-154.
+
Topics to be covered will include these:
 +
* How honeybees learn to navigate using just a few examples
 +
* Why children first learn to cry and then only later to speak
 +
* Rage against the machine: when and why Big Data actually becomes harmful for learning and how Small Data helps
 +
* When Bayes’ Rule fails and when it succeeds
 +
* How children learn rules of language
 +
* Evolution and sex as search
 +
* From birdsong to syntax – the linguistics of birdsong and its relationship to human language
 +
* When neural networks fail
 +
* Google MisTranslate™
 +
* How children learn the meanings of words
 +
* Why only us? – the evolutionary origin of human language
-
Since this add-on is only 3 units, the workload is minimal. Aside from some optional reading between lectures, the only assignment will be a '''five- to eight-page paper''' addressing, researching, or analysing some aspect of human evolution or language that we discuss in class.
+
Requirements will consist of two short, written “reading and response” type assignments along with a slightly longer final written assignment due December 8 on a topic of your choice, subject to instructor approval. There will be no quizzes or final exam.
 +
 
 +
If you have any additional questions, please feel free to email me - otherwise, if you’re interested in joining, or just curious, I’ll see you this coming Wednesday or Friday. I look forward to seeing you!
 +
 
 +
<!--
 +
= Introduction =
 +
 
 +
The 2017 edition of 6.034 offers an optional 3-unit add-on set of lectures given by Professor Robert C. Berwick.
 +
 
 +
This 3-unit section, 6.S077, supplements the main lectures and recitations, focusing on the ''science'' side of 6.034. In particular, 6.S063 addresses long-standing scientific questions and biological intelligence, rather than existing tools for building applications. The topics of '''evolution''' and '''human language''' are heavily emphasized.
 +
-->
<!--
<!--

Revision as of 00:51, 7 September 2017

Prospectus

This term we will again offer an optional 3-unit undergraduate add-on to 6.034, 6.S077, taught by Prof. Robert C. Berwick. It will focus on scientific questions in the field and biological intelligence: what makes animals, including us, smart? How do we and animals learn? There will be a special emphasis on language.

Lectures will be Wednesdays 3-4 pm in room 36-153 and then repeated Fridays 2-3 pm in room 36-155; you can go to whichever lecture works for you. Additional machinery for signing up will be posted soon on the main course Wiki. There will be a lecture the first week of class on both Wednesday and Friday to introduce the add-on so that you can decide whether or not you want to pursue it.

This year, one overarching theme will be how animals, including us, use “Small Data” rather than “Big Data” for learning, along with built-in constraints evolution has provided.

Topics to be covered will include these:

  • How honeybees learn to navigate using just a few examples
  • Why children first learn to cry and then only later to speak
  • Rage against the machine: when and why Big Data actually becomes harmful for learning and how Small Data helps
  • When Bayes’ Rule fails and when it succeeds
  • How children learn rules of language
  • Evolution and sex as search
  • From birdsong to syntax – the linguistics of birdsong and its relationship to human language
  • When neural networks fail
  • Google MisTranslate™
  • How children learn the meanings of words
  • Why only us? – the evolutionary origin of human language

Requirements will consist of two short, written “reading and response” type assignments along with a slightly longer final written assignment due December 8 on a topic of your choice, subject to instructor approval. There will be no quizzes or final exam.

If you have any additional questions, please feel free to email me - otherwise, if you’re interested in joining, or just curious, I’ll see you this coming Wednesday or Friday. I look forward to seeing you!


Personal tools