Decide our topic

[stevengogogo]

We will need to decide a proposal for the incoming final project.

Project Repo

• Those Tex files can be edited on overleaf (find the link in ReadME)
• We need to work on doc/proposal.tex first

Decision 🤝

Dataset

• https://www.kaggle.com/competitions/house-prices-advanced-regression-techniques/data

Framework

• Bayesian learning: https://github.com/stevengogogo/ECEN649_FinalProject/issues/3#issuecomment-1301164220

Format

The proposal should contain: • Project title and team. • Project description. Describe briefly the goal of the project, the data set, and the pattern recognition techniques to be used (e.g., data cleaning, data visualization/exploration, feature selection/extraction, classification/regression method, model selection, error estimation). • References: these are the papers where the experiments were originally published or that support the methodology you will use in your paper. At least 2 papers should be included.

The project must use pattern recognition and machine learning methods. It cannot be based only on simulated data (but it may use it). The results in the project must be your own original work.

You may use a data set that has been published in a recent paper, a data set you produced in your research, or one of the data sets in the textbook (please see Appendix A8).

Platforms to find

• SOTA https://paperswithcode.com/sota
• Kaggle
• Medical Segmentaion Decathlon

References

• Class_Project (1).pdf

[stevengogogo]

Additional comment from Ulisses

For the class project, you can propose classification or regression topics. It does not need to be based on classification only.

[stevengogogo]

[Proposal] Medical Decathlon

http://medicaldecathlon.com/

Declare the current challenges of each dataset.

• All segmentation problem
  • target heretorgenity
  • Small target/ large image

Data set

contains all kinds of organs. They have their own difficulty

[stevengogogo]

[Proposal] Kaggle House pricing prediction -- Regression

https://www.kaggle.com/competitions/house-prices-advanced-regression-techniques/data

• Data
  • CSV with all kinds of property information

[Proposal] Avocado prices -- Regression

https://www.kaggle.com/datasets/neuromusic/avocado-prices

• Data
  • CSV
  • Historical data

[Proposal] Breast cancer mortality

https://www.kaggle.com/datasets/mpwolke/cusersmarildownloadsbreastccsv

[Proposal] Recommend e-commerce

The goal of this competition is to predict e-commerce clicks, cart additions, and orders. You'll build a multi-objective recommender system based on previous events in a user session.The training data contains full e-commerce session information. For each session in the test data, your task it to predict the aid values for each session type thats occur after the last timestamp ts in the test session. In other words, the test data contains sessions truncated by timestamp, and you are to predict what occurs after the point of truncation.

https://www.kaggle.com/competitions/otto-recommender-system/data

[stevengogogo]

COVID

Vietnamese COVID dataset

https://github.com/VinAIResearch/PhoNER_COVID19

Kaggle COVID datasets

https://www.kaggle.com/datasets?topic=covidDataset

[VillaHsu]

Kaggle COVID datasets

Deaths By Risk Factor Dataset

https://www.kaggle.com/datasets/ahmadjalalmasood123/deaths-by-risk-factor-dataset

[stevengogogo]

Framework

Bayesian Neural Netwok

https://jamesmccaffrey.wordpress.com/2021/08/30/an-example-of-a-bayesian-neural-network-using-pytorch/

Paper

Bayesian Deep Learning and a Probabilistic Perspective of Generalization https://arxiv.org/abs/2002.08791

Tutorial for Bayesian learning

• Julia: https://turing.ml/v0.21/tutorials/13-seasonal-time-series/
• Tensorflow: https://www.tensorflow.org/probability/examples/Bayesian_Gaussian_Mixture_Model
• Pytorch: https://jamesmccaffrey.wordpress.com/2021/08/30/an-example-of-a-bayesian-neural-network-using-pytorch/