Laptop Price Estimator

This project aims to predict the prices of laptops using machine learning techniques. The dataset required extensive cleaning and feature engineering due to variations in column values.

Streamlit Application

Project Overview

The dataset consists of information about various laptops, including their specifications and prices. Each entry contains details such as brand, type, screen size, processor, RAM, storage, graphics card, operating system, weight, and price.

Data Cleaning and Feature Engineering

The dataset underwent significant cleaning and feature engineering to handle inconsistencies and prepare it for modeling. This process involved standardizing values, handling missing data, and transforming features to improve predictive performance.

Dataset Overview

The following table provides an overview of the dataset:

ID	Brand	Type	Screen Size (inches)	Display	Processor	RAM	Storage	Graphics Card	Operating System	Weight (kg)	Price (USD)
11	HP	Notebook	15.6	Full HD 1920x1080	Intel Core i3 6006U 2GHz	4GB	500GB HDD	Intel HD Graphics 520	No OS	1.86	18381.0672
12	Apple	Ultrabook	15.4	IPS Panel Retina Display 2880x1800	Intel Core i7 2.8GHz	16GB	256GB SSD	AMD Radeon Pro 555	macOS	1.83	130001.6016
13	Dell	Notebook	15.6	Full HD 1920x1080	Intel Core i3 6006U 2GHz	4GB	256GB SSD	AMD Radeon R5 M430	Windows 10	2.2	26581.392

Machine Learning Models

Multiple machine learning models were experimented with to predict laptop prices, including:

• Linear Regression
• Ridge Regression
• Lasso Regression
• Support Vector Machines (SVM)
• Random Forest
• Decision Trees Regressor
• KNeighborsRegressor
• Adaboost
• XGBoost
• Gradient Boost
• Extra Tree
• Voting Classifiers with different models
• Stacking method with different models

Best Performing Model

After evaluating various models, the Random Forest algorithm yielded the best results in predicting laptop prices.

Streamlit Application

The project includes a Streamlit application deployed at Laptop Price Estimator. Users can interact with the application to estimate laptop prices based on their specifications.

Repository Contents

• Dataset: Contains the cleaned dataset used for training and evaluation.
• Notebooks: Jupyter notebooks detailing the data cleaning, feature engineering, and model training processes.
• Models: Includes trained models for predicting laptop prices. pipe.pkl utilizes the Random Forest algorithm for predictions.

Usage

To replicate the project or explore further:

1. Clone this repository.
2. Install the necessary dependencies specified in requirements.txt.
3. Explore the notebooks to understand the data preprocessing and modeling steps.
4. Utilize the provided models for predicting laptop prices.
5. Interact with the Streamlit application for estimating laptop prices based on custom specifications.