farah-dhiaf / FP-Data-Mining

Final Project for Data Mining: Flood Probability Forecasting

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

FP-Data-Mining

Nama Anggota NRP
Farah Dhia Fadhila 5025211030
Salsabila Fatma Aripa 5025211057
Katarina Inezita 5025211148
Alfa Fakhur Rizal Zaini 5025211214

Prediction of Flood Probability using regression and ensemble method. The dataset used is from Kaggle Competition (https://www.kaggle.com/competitions/playground-series-s4e5) and the goal is to predict the probability of a region flooding based on various factors.

There are 4 scenarios:

  1. Scenario 1: Remove outliers + Standard Scaler
  2. Scenario 2: Remove outliers + MinMax Scaler
  3. Scenario 3: No remove outliers + Standar Scaler
  4. Scenario 4: No remove outliers + MinMax Scaler

Those scenarios will be implemented into 4 regression method:

  1. Linear Regression
  2. Logistic Regression
  3. XGBoost
  4. Light GBM

The models will be evaluated based on MSE, RMSE, and R2.

About

Final Project for Data Mining: Flood Probability Forecasting