Basic Information

• Instructor: Jie Wang
• Email: jiewangx@ustc.edu.cn
• Time and Location: Tues., Thur. 14:00 PM - 15:35 PM (3C304)
• TAs:
  • Xize Liang (xizeliang@miralab.ai)
  • Zijie Geng (zijiegeng@miralab.ai)
  • Haoyang Liu (haoyangliu@miralab.ai)

Lectures

All course materials will be shared via this page.

Project

Description

• Recommending clothes of suitable sizes to customers based on the information of clothes and users are very important for E-commerce platforms. In this project, you are expected to implement a classifier to predict customers’ fit feedback (“Large”, “True to Size”, or “Small”) based on a dataset collected from RentTheRunWay.

Dataset

• You can download the training data (in the form of json data) from here (Updated November 24, 2022). The dataset contains 87766 samples. Each sample is a dictionary corresponding to a rental or purchase record as shown in the following example.

  Key	Value
  fit	True to Size
  item_name	Lilly Pulitzer Lettie Eyelet Top
  size	XS
  price	$158
  user_name	Alexandra
  rented_for	Everyday
  usually_wear	4
  age	58
  height	5’ 2''
  weight	116LBS
  body_type	PEAR
  bust_size	32B
  review_summary	Unexpected wow
  review	Thought this would be a whatever shirt. Really cute!! Color popped and fit well. Definitely rent.
  rating	5

• Notice: For some training samples, some features and even labels are missing. Before building a model, you may want to clean the data or think about how to use the data whose features or labels are missing.

• Your model will be tested on a dataset (in the form of json, too) of 9751 samples. In the testing dataset, labels “Small”, “True to Size”, and “Large” are set to 1, 2, and 3, respectively. Each testing sample is a dictionary as follows.

  Key	Value
  fit	1
  item_name	Amanda Uprichard Tweed Aldridge Blazer
  size	S
  price	$319
  rented_for	Work
  usually_wear	4
  age	40
  height	5’ 3''
  weight	130LBS
  body_type	STRAIGHT & NARROW
  bust_size	34B

Requirements

• Do NOT use any autograd tool or any optimization tool from machine learning packages. You are supposed to implement your algorithm from scratch. For example, if you want to use a neural network, you are expected to implement both forward and backward passes. You can use the packages in the WhiteList. TAs will update the Whitelist if your requirements are reasonable.

• You can work as a team with no more than three members in total. Please list the percentage of each member’s contribution in your report, e.g., {San Zhang: 30%, Si Li: 35%, Wu Wang: 35%}.

• We define a default base class called PB20000000 in PB20000000.py. You are supposed to implement your algorithm in [your student ID] directory. We provide an example code here. For detailed requirements, please refer to the comments in our code.

• You are supposed to send a package named [your student ID].zip , which contains the [your student ID] directory organized as follows to ml2022fall_ustc@163.com.

  [your student ID]
    |- [your student ID].py 
    |- [your student ID]-report.pdf (your report)
    |- ... (your code and model)
  

  For a teamwork, please use the team leader’s student ID in the package name and submit the package by your team leader.

• Remember to save the trained model. You are supposed to send your trained model to the aforementioned e-mail address.

• Please submit a detailed report. The report should include all the details of your projects, e.g., the implementations, the experimental settings and the analysis of your results.

Grading

• The full points = min(Base score (up to 20pts) + Bonus (up to 5pts), 20pts).
• The base score is determined by the Macro F1-score evaluated by TAs’ code.
• The bonus involves three aspects as follows.
  • Your insights on the data and task.
  • The novelty of your approach, which should be highlighted in your report.
  • The readability of your code and report. Please make them easy to follow.

System Requirements

• We will evaluate your model on a GeForce RTX 2080Ti (about 10G memory) under Ubuntu 18.04 system. Please limit the size of your model to avoid OOM.

Hint

• These papers may be helpful to you.

  [1] Yoon Kim. 2014. Convolutional Neural Networks for Sentence Classification. In Proceedings of EMNLP 2014.

  [2] Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. Bert: Pre-training of deep bidirectional transformers for language understanding. In Proceedings of NAACL 2019.

Due Day

• Team leaders should inform the TAs about your team members before 23:59 PM, December 2, 2022.

• Please submit your report, code and trained model before 23:59 PM, January 11, 2023.

• No late submissions will be accepted.

---

Page view (from Jan 1, 2021)：

Downloads (from Jan 1, 2021)：112