# Collaborative learning confusion

**URL:** <https://community.deeplearning.ai/t/collaborative-learning-confusion/510049>\
**Category:** Unsupervised Learning, Recommenders, Reinforcement\
**Tags:** week-module-2\
**Created:** [December 3, 2023, 9:21pm UTC](https://community.deeplearning.ai/t/collaborative-learning-confusion/510049 "2023-12-03T21:21:58Z")\
**Posts on this page:** 5\
**Page:** 1

<div class="post-metadata">

**Author:** ![taha\_fawzy\_anwar](https://avatars.discourse-cdn.com/v4/letter/t/41988e/32.png) [@taha\_fawzy\_anwar](https://community.deeplearning.ai/u/taha_fawzy_anwar)\
**Post date:** [December 3, 2023, 9:21pm UTC](https://community.deeplearning.ai/t/collaborative-learning-confusion/510049/1 "2023-12-03T21:21:58Z")

</div>

hello ,in collaborative learning assigment 1 ,why did I retrained whole parameters after adding new user column ,what I only need weights of new user ,have i misunderstanded anything

---

<div class="post-metadata">

**Author:** ![TMosh](https://sea1.discourse-cdn.com/dlai/user_avatar/community.deeplearning.ai/tmosh/32/9321_2.png) [@TMosh](https://community.deeplearning.ai/u/TMosh)\
**Post date:** [December 3, 2023, 10:15pm UTC](https://community.deeplearning.ai/t/collaborative-learning-confusion/510049/2 "2023-12-03T22:15:11Z")

</div>

When you add a user, you need to re-train so their review can be included in the weights for making new predictions.

---

<div class="post-metadata">

**Author:** ![taha\_fawzy\_anwar](https://avatars.discourse-cdn.com/v4/letter/t/41988e/32.png) [@taha\_fawzy\_anwar](https://community.deeplearning.ai/u/taha_fawzy_anwar)\
**Post date:** [December 3, 2023, 10:27pm UTC](https://community.deeplearning.ai/t/collaborative-learning-confusion/510049/3 "2023-12-03T22:27:36Z")

</div>

okay ,why don’t i used anyway to train weights only for the new user ,to reduce computationality ,save last info?

---

<div class="post-metadata">

**Author:** ![TMosh](https://sea1.discourse-cdn.com/dlai/user_avatar/community.deeplearning.ai/tmosh/32/9321_2.png) [@TMosh](https://community.deeplearning.ai/u/TMosh)\
**Post date:** [December 3, 2023, 11:04pm UTC](https://community.deeplearning.ai/t/collaborative-learning-confusion/510049/4 "2023-12-03T23:04:57Z")

</div>

You get the best performance if all of the weights are computed including all of the users.

---

<div class="post-metadata">

**Author:** ![taha\_fawzy\_anwar](https://avatars.discourse-cdn.com/v4/letter/t/41988e/32.png) [@taha\_fawzy\_anwar](https://community.deeplearning.ai/u/taha_fawzy_anwar)\
**Post date:** [December 3, 2023, 11:06pm UTC](https://community.deeplearning.ai/t/collaborative-learning-confusion/510049/5 "2023-12-03T23:06:41Z")

</div>

ok thank you