Bulldozer Sale Price Prediction

🚜 Bulldozer Sale Price Prediction
Project Type: Regression Model | Industry Focus: Heavy Equipment / Construction | Tools Used: Python, Scikit-learn, Pandas, Matplotlib, XGBoost
🔍 Project Overview
This project aims to predict the sale price of used bulldozers using historical data and machine learning regression techniques. Accurate price forecasting helps construction firms and dealers make informed purchasing and resale decisions.
🎯 Objectives
- Develop a regression model to estimate the resale price of bulldozers
- Analyze the impact of machine specifications, usage, and year of manufacture on price
- Assist stakeholders in asset valuation and financial planning
🧩 Dataset
- Source: Blue Book for Bulldozer Sales (via Kaggle)
- Features: Model ID, Year Made, Machine Hours, Product Size, Usage Type, Enclosure, Hydraulics, etc.
- Target: Sale Price (USD)
🛠️ Techniques & Methodology
- Data cleaning: handled null values, parsed date features, and removed outliers
- Feature engineering: extracted year sold, equipment age, and categorical encodings
- Applied models: Linear Regression, Random Forest Regressor, XGBoost Regressor
- Evaluated using RMSLE (Root Mean Squared Log Error), R² score
🚀 Key Results
- XGBoost Regressor delivered the best performance with RMSLE under 0.25
- Model year, usage hours, and product size were strong predictors of price
- Final model generalized well on unseen data with low prediction error
💼 Business Impact
This pricing model can be deployed by equipment dealers and auction platforms to offer real-time price estimates, identify undervalued assets, and streamline procurement strategy—reducing overpayment risk in multi-million dollar equipment portfolios.
Project Overview
Customers who left within the last month - the column is called Churn Services that each customer has signed up for - phone, multiple lines, internet, online security, online backup, device protection, tech support, and streaming TV and movies Customer account information - how long they've been a-customer, contract, payment method, paperless billing, monthly charges, and total charges Demographic info about customers - gender, age range, and if they have partners and dependents Inspiration To explore this type of models and learn more about the subject.
Code Snippet
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
from sklearn.impute import SimpleImputer
from sklearn.model_selection import train_test_split
from sklearn.ensemble import RandomForestRegressor
from sklearn.metrics import (r2_score,
mean_absolute_error,
mean_squared_error)
import pickle
sns.set()
import os
image_path=os.path.join('.','images','all_image')
os.makedirs(image_path,exist_ok=True)
def savefig(fig_id,tight_layout=True,resolution=300,fig_extension='png'):
path=os.path.join(image_path,fig_id+'.'+fig_extension)
print(f'saving figure {fig_id}')
if tight_layout:
plt.tight_layout()
plt.savefig(path,format=fig_extension,dpi=resolution)



















