Instructor

• Virgil Pavlu — v.pavlu@northeastern.edu (prefers Teams chat)
• Lectures: Mon & Thu 11:45am–1:25pm. Room SH 335

TAs

• Bhanu Sharma, sharma.bhan@northeastern.edu
• Danush Gopinath, gopinath.da@northeastern.edu
• Rohith Kumar Senthil, senthilkumar.ro@northeastern.edu
• Sankar Sudheer Ayachitula, ayachitula.s@northeastern.edu

Helping Sessions / Office Hours

• Boston Campus: Mon 6–7:30pm (HA-321), Wed 6–7:30pm (RY-431), Thu 7:30–9pm (HS-107)

• TA/OH Tentative Schedule Fall 26 CS6140

  	MON	TUE	WED	THU	FRI	SAT	SUN
  9
  9:30
  10
  10:30
  11		Rohith	Sankar		Rohith
  11:30	CLASS	Rohith	Sankar	CLASS	Rohith
  12	CLASS	Rohith	Sankar	CLASS	Rohith	Bhanu	Bhanu
  12:30	CLASS	Rohith	Sankar	CLASS	Rohith	Bhanu	Bhanu
  1	CLASS		Danush	CLASS		Bhanu	Bhanu
  1:30			Danush			Bhanu	Bhanu
  2			Danush
  2:30			Danush
  3		Danush
  3:30		Danush
  4		Danush			Bhanu
  4:30		Danush			Bhanu
  5					Bhanu
  5:30					Bhanu
  6	HA321 Rohith		RY431 Danush
  6:30	HA321 Rohith		RY431 Danush
  7	HA321 Rohith		RY431 Danush
  7:30				HS107 Sankar
  8				HS107 Sankar
  8:30				HS107 Sankar
  9

OH format: queue, in order of arrival. TAs might leave after 1 hour if no students are in the queue waiting.

Teams: Khoury - CS 6140 3 (Fall 2026) — not required, but some OH/recitations, code demos, or even lectures might happen online if necessary.

Discussion

Piazza

Please use the discussion forum for all questions regarding material, assignments, due dates, data issues, programming issues, etc. Do not email TAs or the instructor for these questions. Personal/private matters (availability, delays, grades, term projects, or other advanced material) should be discussed by email.

Recommended Texts (not mandatory)

• [CB] Pattern Classification (2nd Edition) by Richard O. Duda, Peter E. Hart, and David G. Stork
• [KMPP] Kevin Murphy: Machine Learning, A Probabilistic Perspective
• [HTP] Hastie, Tibshirani: An Introduction to Statistical Learning: with Applications in Python
• [ZLLS] Dive into Deep Learning by Aston Zhang, Zachary C. Lipton, Mu Li, Alexander J. Smola

Other texts

• Pattern Recognition and Machine Learning by C. Bishop
• Introduction in Machine Learning by Ethem Alpaydin
• Machine Learning by Tom Mitchell
• Machine Learning in Action by Peter Harrington

Prerequisites

• Linear algebra tutorial: stanford.edu/class/cs229/section/cs229-linalg.pdf
• Probabilities tutorial: stanford.edu/class/cs229/section/cs229-prob.pdf
• Convexity tutorial: part 1, part 2
• Programming: you should be familiar with programming in an imperative language like Java, Matlab, or Python. While the programming language is your choice, we strongly encourage one of these three. If you choose not to use one of these, discuss with the instructor first.
• CS5800 or similar course in Algorithms or problem solving

Homework

• There will be about seven assignments (one per module, every two weeks), including programming; see the syllabus for an exact schedule. Some problems can be difficult, and it will often help to discuss them with others — feel free to form study groups. However, the idea is for everyone to understand the problems and experience working through the solutions, so you may not simply "give" a solution to another classmate. Each student must write up their own homework solutions/code and must not read or copy the solutions/code of others. If you work with others on a problem, you must note with whom you discussed the problem at the beginning of your solution write-up.
• We expect the average total load to be less than 20 hours/week. Students can spend more than 20 hours/week on additional work denoted "Extra Credit" or "Optional, no credit" — this is not for grade points: harder, more time consuming, no points, and less discussed during helping sessions; it's for students who want to get more out of the course.
• "GR_ONLY" exercises are required for CS6140 (grad level) but not required for 4420 (UG level).
• Homework is generally due on Tuesdays, submitted via Gradescope; you may demo the assignment to a TA for grading up to 7 days after the deadline.

Late homework policy: homework is due on the announced due date. You will be granted one homework extension of 7 days total, to be used at your discretion, no questions asked (this policy does not apply to projects). After the first 7 late days, unexcused late assignments will be penalized 10% per calendar day late, but no more than 60% of the total (i.e. you can still get 40% of the points by submitting before the term ends).

Projects

• Optionally students can work on a project, if they discuss it in advance with the instructor.

Final Exam

• We plan to have a final exam, 3 hours. The exam is theory, on paper, in class: open book, open notes, open HWs — all must be printed or handwritten on paper (no devices with internet access). Do not expect pure math derivations/calculations — expect questions on key algorithmic points related to class material, and possible simple modifications for a specific purpose. See the Exam page for details.

Academic Honesty

• All work submitted for credit must be your own.
• You may discuss the homework problems or projects with your classmates, the teaching assistant(s), and instructor. You must acknowledge the people with whom you discussed your work, and you must write up your own solutions. Any written sources used (apart from the text) must also be acknowledged; however, you may not consult any solutions from previous years' assignments whether they are student or faculty generated.

LLM / AI Usage

• Exam: no LLM (no internet-connected devices allowed at all).
• Homework — coding: LLM is fine for boilerplate, debugging, plotting, and general questions. The specific lines marked TODO_STUDENT in each notebook are the actual learning objective — attempt those yourself first; an LLM can help you check or debug your own attempt afterward, but shouldn't be the one producing that step for you.
• Homework — theory: don't use an LLM to get the answer directly. It's fine to use one to polish your writing, run a validation experiment, or connect your answer to the lecture/notes.
• Project: LLM does essentially everything (coding, execution, debugging, plots, math) except the report, which you write yourself. See the Project page.