News
Assignment #9 UpdatedWritten on 13.06.25 by Monseej Purkayastha Hi everyone, There was a typo in the assignment pdf, which has been corrected now. Please use the updated version for your assignment. Best regards, |
Assignment #9 ReleasedWritten on 11.06.25 by Monseej Purkayastha Hi everyone, Assignment #9 has been released. The deadline for this assignment is June 18, 2025 at 23:59.
Best Regards, Your SNLP Tutors |
Assignment #8 UpdatedWritten on 09.06.25 by Oliver Bintriem Hi everyone, Note that Assignment 8 got updated, as for exercise 4.2 the wrong graphs were attached with the assignment. Please take a look at this exercise again and consider the new graphs. Best regards, Your SNLP Tutors |
Assignment #8 ReleasedWritten on 04.06.25 by Oliver Bintriem
Hi everyone, Assignment #8 has been released. The deadline for this assignment is June 11, 2025 at 23:59.
Best Regards, Your SNLP Tutors |
Assignment #7: Modified + Deadline ExtendedWritten on 03.06.25 by Mikhail Sonkin Hi everyone, Due to the difficulties you've expressed with the latest assignment, we have uploaded another version of it. Note that the only differences are: 1. The sizes of the datasets (500 & 300 entries) 2. The uni- and bigram in Exercise 1 The finance dataset remains the same size,… Read more Hi everyone, Due to the difficulties you've expressed with the latest assignment, we have uploaded another version of it. Note that the only differences are: 1. The sizes of the datasets (500 & 300 entries) 2. The uni- and bigram in Exercise 1 The finance dataset remains the same size, because the bonus exercise calls for more resources. Because of the late adjustment, the new deadline for Assignment 7 is Friday, 06.06.2025, 19:00. Have a nice week, Your SNLP Tutors |
Assignment #7: Large DatasetsWritten on 03.06.25 by Mikhail Sonkin Dear students, it seems like for some of you the smaller datasets still yielded memory errors. I apologise if this has caused to hinder your progress with the assignment. This is to notify that you can cut the provided datasets to create smaller matrices. As long as the underlying code is… Read more Dear students, it seems like for some of you the smaller datasets still yielded memory errors. I apologise if this has caused to hinder your progress with the assignment. This is to notify that you can cut the provided datasets to create smaller matrices. As long as the underlying code is correct, we will grant you the full points, no matter what the data size is. Best regards, Mikhail Sonkin |
Assignment #7 ModifiedWritten on 30.05.25 by Mikhail Sonkin Dear all, please note that Assignment #7 was just updated: in Exercise 2 you need to train a decision tree classifier instead of Random Forest. Also, it's been mentioned that some of you weren't able to complete the assignments due to the large datasets. The newly published assignment has… Read more Dear all, please note that Assignment #7 was just updated: in Exercise 2 you need to train a decision tree classifier instead of Random Forest. Also, it's been mentioned that some of you weren't able to complete the assignments due to the large datasets. The newly published assignment has smaller samples of the same datasets: feel free to use those. If there are any other issues you run into, please ask them on the forum! Have a nice weekend, Your SNLP Tutors |
Assignment #7 ReleasedWritten on 28.05.25 by Mikhail Sonkin Hi everyone, Assignment #7 has been released. The deadline for this assignment is June 4, 2025 at 23:59.
Best Regards, Your SNLP Tutors |
Assignment #6 UpdatedWritten on 25.05.25 by Monseej Purkayastha Hi everyone, We wanted to inform you about a small correction in the notebook for Exercise 6.3 in Assignment #6. Initially, the notebook code was only performing standard 1-fold cross-validation instead of the intended 5-fold cross-validation. This issue has now been fixed. Important: Your task… Read more Hi everyone, We wanted to inform you about a small correction in the notebook for Exercise 6.3 in Assignment #6. Initially, the notebook code was only performing standard 1-fold cross-validation instead of the intended 5-fold cross-validation. This issue has now been fixed. Important: Your task remains unchanged — you only need to complete the functions in exercise_3.py. The updated code will now correctly handle 5-fold CV on its own. If you’ve already modified the driver code to implement 5-fold CV yourself, that’s absolutely fine. In that case, please just leave a short comment indicating the lines where you made changes.
Best regards, |
Assignment #6 ReleasedWritten on 21.05.25 by Monseej Purkayastha Hi everyone, Assignment #6 has been released. The deadline for this assignment is May 28, 2025 at 23:59.
Best Regards, Your SNLP Tutors |
Assignment #5 ReleasedWritten on 14.05.25 by Maham Shakir Hi everyone, Assignment #5 has been released. The deadline for this assignment is May 21, 2025 at 23:59.
Best Regards, Your SNLP Tutors |
Assignment #4 ReleasedWritten on 07.05.25 by Oliver Bintriem Hi everyone, Assignment #4 has been released. The deadline for this assignment is May 14, 2025 at 23:59.
Best Regards, Your SNLP Tutors |
Assignment #3 ReleasedWritten on 30.04.25 by Monseej Purkayastha Hi All, Assignment #3 has been released. The deadline for this assignment is May 7, 2025 at 23:59.
Best Regards, Your SNLP Tutors |
Statistical Natural Language Processing
Location: E13, HS 1
Time: Wednesday 14:15-15:45
Starts: April 16th
Suitable for: CS, DSAI, CuK, ES, CoLi, Visual Computing, 6CP
- Introduction
- Natural Language as a Sequence of Symbols
- Basics of Language Modeling
- Entropy
- Backing-Off Language Modeling
- Text Classification
- Word Sense Disambiguation
- CRFs and Sequence Labeling
- Information Retrieval
- Machine Translation