HANDS-ON-LAB

Predicting Price of a Refurbished Mobile and Tablet Device

Problem Statement

The used and refurbished device market has grown considerably over the past decade as it provides cost-effective alternatives to consumers and businesses looking to save money when purchasing one. Maximizing the longevity of devices through second-hand trade also reduces their environmental impact and helps recycle and reduce waste. Here is a sample dataset of normalized used and new pricing data of refurbished/used devices.

Dataset

The data contains 15 variables and of which normalized_used_price is the Target variable.

The complete data dictionary can be found here.

Kindly download the data from here.

Tasks

Learn about and Install the package duckdb for using sql queries on the dataframe
Hypothesis-based EDA:

- Which brand and OS have higher resale value?
- Create a regplot with days_used and normalized_used_price to understand the relationship between them.
- Does the average phone resale value decrease in recent years because of the availability of mobiles in all different price ranges? (plot trend of average resale price for each year)

Create a correlation heatmap between the numerical and target variables to understand the correlation of different features with the target.
Analyze the categorical variables using Chi-squared, Cramer's V, and ANOVA tests.
Build MARS and GAM models on the final data. Compare the results.

Build advanced predictive models and compare the results using MARS and GAM techniques.

FAQs

Q1. Which brand and operating system have a higher resale value?

The analysis will provide insights into brands and OS with higher resale values based on the dataset.

Q2. What is the relationship between days_used and normalized_used_price?

By creating a regplot, the analysis will reveal the relationship between days_used and normalized_used_price.

Q3. Has the average phone resale value decreased in recent years due to availability in different price ranges?

The analysis will plot the trend of average resale prices for each year, providing insights into any potential decrease.