search

AtomAnu / COMP0036_EPL_Prediction

0 stars 0 forks source link

Algorithm ideas #4

Open AtomAnu opened 4 years ago

AtomAnu commented 4 years ago

Explain your approach down below.

gbifvyihk commented 4 years ago
  1. Separate data into different teams
  2. Create a logistic regression model (maybe for different data) for each team
  3. Use cross validation to train the model
  4. Add weight between different data to predict the final result (optimized by model selection)
AtomAnu commented 4 years ago

Hey all, I will conclude and post my idea in a few hours. Meanwhile, if you have not seen it yet, this paper could be useful: https://www.imperial.ac.uk/media/imperial-college/faculty-of-engineering/computing/public/1718-ug-projects/Corentin-Herbinet-Using-Machine-Learning-techniques-to-predict-the-outcome-of-profressional-football-matches.pdf.

Also, @gbifvyihk do you agree that we should create other models as well as the logistic regression model and select the best one for each feature?