In this notebook, I have made an attempt to develop a model to solve the Kaggle Problem Statement - Flight Fare Prediction using ML Regression. Here, I have implemented various ML regression models and finally decided upon Random Forest Regressor for predictions.
What I do in this project a) Data collection from Kaggle b) Perform Data Cleaning / Data Preparation / Data Pre-processing c) Data visuaslisation(EDA) d) Perform feature engineering I) Feature encoding II) checking outliers & impute it..
e) build machine leaning model & dump it..
Flight Fare Prediction using ML Regression
• The dataset for the project is taken from Kaggle, and it is a time-stamped dataset so, while building the model, extensive pre-processing was done on the dataset especially on the date-time columns to finally come up with a ML model which could effectively predict airline fares across various Indian Cities
Life Cycle of any ML Project!
-
Problem Solving a) North Star Metrics (define success) b) Build a rough sketch for the problem
-
Data Collection a) Frame the requirement b) Define quantity & quality
-
EDA a) Data Cleaning b) Data QC c) Feature Engg
-
Data Selection a) Split (Train Vs Test Vs Val) b) Random/ Stratified
-
Pre - Processing
-
Train the model a) Baseline model b) Now build on top (hyper parameter) c) Record your metrics d) Save the model
-
Validation
-
Testing
-
Inferencing a) Build a system that will help the other application to use model