Logo Lanfrica

aditya1876/AlgerianFirePrediction

Domain:

environment and energy

Record type:

dataset
Creator:
adi
Host:
Predict the temperature and Fire based on data from Algerian Forests # AlgerianFirePrediction Predict the temperature and Fire based on data from Algerian Forests ##### Dataset - archive.ics.uci.edu ### Objectives - import data - EDA: - create detailed profile - perform graph based analysis - statistical insite from data - preporcessing - handle missing values - encode categorical data - scale dataset if variation is high - handle outlier - handle multicolinearity - Create classification and regression model - Regression model: - linear, ridge, lasso, SVR, Decision Tree Regressor, Random Forest Regressor - Cross validation and hyperparameter tuning - mse for each model - select best model based on R2 score - Classification model: - Logistic, SVM, DecisionTreeClassifier, naive bayes, Random forest classifier - crossvalidation and hyperparameter tuning - Classification report and select best model on the basis of the report - Generic Tasks: - Create Flask API(postman) and HTML page - Perform single value prediction and bulk prediction - Data load via mongo DB or mysql(during bulk prediction) - Modular code - logging - handle exception at every step #### Notes: - conda env - proj_algfire