NAME: Muhamamd Tuaha

ID:225176
SUBMITTED TO: Sir Ghulam Ali
Full Project.
Table of Contents
Title: Smartphone Price Prediction using Data Science.........................................................................2
1. Introduction.......................................................................................................................................2
2. Problem Statement............................................................................................................................2
3. Datasets..............................................................................................................................................3
3.1 Source:..........................................................................................................................................3
3.2 Dataset Description:....................................................................................................................3
3.3 Key Attributes:.............................................................................................................................3
4. Analytical Approach..........................................................................................................................3
4.1 Exploratory Data Analysis (EDA):.............................................................................................3
4.2 Modeling:.....................................................................................................................................3
4.3 Evaluation:...................................................................................................................................3
5. Preprocessing Techniques.................................................................................................................4
6. Models................................................................................................................................................4
7. Results................................................................................................................................................4
8. Findings and Conclusions.................................................................................................................4
9. Deployment Strategy.........................................................................................................................5
10. Feedback and its analysis................................................................................................................5
11. Future Goals....................................................................................................................................5

Project Report:

Title: Smartphone Price Prediction using Data Science

The smartphone market is vast and competitive, with prices fluctuating significantly based on
various features. This project aims to build a predictive model to classify smartphones into price
categories based on their technical specifications, helping both consumers and manufacturers
make informed pricing decisions.

2. Problem Statement

The primary problem addressed in this project is:

The smartphone market is filled with various devices offering a wide range of features, making it
difficult for both consumers to find the best value for money and manufacturers to price new
devices correctly. This project solves that problem by using data science techniques to predict
price categories based on device specifications.

3. Datasets
3.1 Source:

The dataset is sourced from Kaggle (Mobile Price Classification Dataset) and is available
publicly.

3.2 Dataset Description:

The dataset contains 2000 rows and 21 columns, with features such as RAM, battery power,
internal storage, screen resolution, camera quality, connectivity options (4G/5G), and processor
speed.

3.3 Key Attributes:

 Battery_power (mAh)
 RAM (MB)
 Internal Memory (GB)
 Screen Resolution (px_height, px_width)
 Processor Speed (GHz)
 Price Range (target variable: 0=budget, 1=mid-range, 2=upper-mid, 3=flagship)

4. Analytical Approach
4.1 Exploratory Data Analysis (EDA):

Was conducted to understand the relationships between features and price ranges. This helped in
feature selection and identifying key attributes like RAM, battery power, and processor speed as
strong price predictors.

4.2 Modeling:

Decision Tree, Random Forest, and Logistic Regression were used to build the predictive model.

4.3 Evaluation:

The models were evaluated using metrics such as accuracy, precision, recall, and F1-score.

5. Preprocessing Techniques

 Normalization of features using StandardScaler.

 Outlier detection was done to identify extreme values (e.g., high-end flagship smartphones).

 Data Splitting: 80% training and 20% testing data split.

6. Models

 Decision Tree Classifier: A simple model used for interpretability.

 Random Forest Classifier: An ensemble method used to improve accuracy and reduce
overfitting.

 Logistic Regression: Used as a baseline model for comparison.

7. Results

 The Random Forest model performed the best, achieving an accuracy of approximately 90%
in predicting smartphone price categories.
 Feature importance analysis revealed that RAM, battery power, and processor speed were
the most significant predictors of price range.

8. Findings and Conclusions

 RAM and battery power are the most influential factors in determining smartphone price.

 The dataset was clean, balanced, and sufficient for building a robust predictive model.

 The use of machine learning algorithms enabled automated, data-driven price categorization.

9. Deployment Strategy

The model was deployed using Streamlit, allowing users to interact with the model and make
predictions on smartphone prices based on provided specifications.

10. Feedback and its analysis

Feedback was collected from users of the deployed application. This feedback highlighted
potential improvements such as adding brand information and improving the UI/UX of the
application.

11. Future Goals

Future improvements include adding features such as brand name, release year, and further
refining the user interface for a better user experience.