RNN Project Information 🠔

• You can get 10 to 20 points

• 📒 Every project must include brief description of the dataset so you get to know the data you are working with

  • Number of instances, number of classes and class balance information
  • Examples of the text data

• 🔎 What model performance metrics have you decided to use?

  • e.g. Accuracy, Precision, Recall, F1-score etc.
  • 📒 State which one of the scores is the most important from your point of view given the class balance, task, ... in a Markdown cell!

• 🎯 Try minimally 2 different models

  • 🚀 The first model should be built from scratch, i.e. create your own architecture and train the model
    • If you perform various experiments or hyper-parameter tuning make sure to include everyting in the Notebook step by step with some brief comments about the experiments
      • 💡 e.g. effect of BatchNormalization,layer sizes or optimizer choice on accuracy,train time, ...
  • 🚀 The second model will use transfer learning techniques
    • You can use any set of pre-trained embedding vectors (GloVe, Word2Vec, FastText etc.) or any transformer-based model (this is optional as it is more advanced approach, see HuggingFace or Keras NLP models)
    • 📌 Fine tune the model for your dataset (if it is possible) and compare it with the models that you created from scratch

General Information

✅ Mandatory part of every project is a summary at the end in which you summarize the most interesting insight obtained.

Upload a 📝 Jupyter Notebook with descriptions included or a PDF report + source codes.

💡 Estimated time for the project is 5-10h, this value heavily depends on your skill, but you can use it as a guidance for a project size.

🎯 Deadline is 26. 5. 2024 🍀

Upload the project to:

• Dropbox:

  • Thu, 09:00: Dropbox
  • Thu, 12:30: Dropbox

• 💡 Dropbox will ask you for your name - use your VSB login please 👍

Example datasets

1 - Sentiment140 dataset with 1.6 million tweets

https://www.kaggle.com/datasets/kazanova/sentiment140

You can select smaller subset of the dataset as it contains pretty high number of tweets.

2 - News Category Dataset

https://www.kaggle.com/datasets/rmisra/news-category-dataset

3 - Coronavirus tweets NLP - Text Classification

https://www.kaggle.com/datasets/datatattle/covid-19-nlp-text-classification

4 - Trip Advisor Hotel Reviews

https://www.kaggle.com/datasets/andrewmvd/trip-advisor-hotel-reviews

Dataset of individual selection

It is highly welcomed if you use a dataset of your individual choice. In this case just drop me an e-mail so I can approve your choice or give some recommendations.

You can find plenty of datasets on Kaggle

If you work with RNN in your semestral project, it is possible to use it instead of this project, even if is not a classification task. Just drop me an e-mail and we will discuss it individually.