MachineLearning

Different DataSet Analysis

Welcome to my Machine Learning project repository! This repo contains beginner-to-intermediate level end-to-end projects using real-world datasets to apply core ML concepts like data cleaning, EDA, preprocessing, model building, evaluation, and visualization.

📊 Projects Included

1. SUV sales analysis

   Goal: Predict whether a user is likely to purchase an SUV based on demographic features like age, salary, and gender.

   Algorithm: Logistic Regression

   Highlights: -Train-Test split and model training using LogisticRegression

   -Evaluation using Accuracy, Confusion Matrix, and Classification Report

   -Visualization of decision boundary using Matplotlib

   DataSet Link: https://www.kaggle.com/datasets/iamaniket/suv-data

2. Student Social media Addiction analysis:

Goal: To predict how social media habits (platforms, frequency, etc.) influence students' academic performance based on self-reported data.

Algorithm: Linear Regression

Highlights:

- One-Hot Encoding of categorical features

- Evaluation using R², RMSE, MAE

- Actual vs Predicted scatter plot

DataSet Link: https://www.kaggle.com/datasets/adilshamim8/social-media-addiction-vs-relationships

3. Titanic Data Analysis:

Goal: This project explores the famous Titanic dataset to identify patterns and insights into passenger survival.

Algorithm: Logistic Regression

Highlights:

• Data cleaning, Handling missing values

• Key Insights using Visualisation

• Survival Analysis

DataSet Link: https://www.kaggle.com/competitions/titanic/data

4. Telecom Customer Data:

Goal: This project analyzes telecom customer data to predict churn (i.e., whether a customer will leave the service provider). Using logistic regression, we build a simple but effective classification model after preprocessing and scaling the data.

Algorithm: Logistic Regression

Highlights:

- Data Preprocessing

- One-Hot Encoding of categorical features

- Evaluation using Accuracy,Precision,Recall,F1 score

- Confusion Matrix

DataSet Link: https://www.kaggle.com/datasets/blastchar/telco-customer-churn

5. Insurance Charges Prediction (Linear vs Ridge vs Lasso Regression)

Goal:To predict medical insurance charges based on personal attributes like age, BMI, region, smoking habits, and more.

Algorithm:Linear Regression, Ridge Regression, Lasso Regression

Highlights:

-Data scaling using StandardScaler before applying Ridge/Lasso

 -Actual vs Predicted scatter plot comparison for all three models

DataSet Link: https://www.kaggle.com/datasets/mirichoi0218/insurance

6. Loan Approval Prediction (Logistic vs SVM vs KNN vs Random Forest)

Goal:To predict whether a loan will be approved based on applicant information such as gender, marital status, income, education, employment status, and property area..

Algorithm:Logistic Regression, Support Vector Machine (SVM), K-Nearest Neighbors (KNN), Random Forest

Highlights:

  -Missing value imputation using mode (for categorical) and median (for numerical)

  -Label Encoding for categorical variables
  
  -Trained and compared four classification models
  
  -Achieved best accuracy of 81% using Random Forest
  
  -Evaluated models using accuracy, confusion matrix, and classification report

DataSet Link: https://www.kaggle.com/datasets/krishnaraj30/finance-loan-approval-prediction-data?select=train.csv

                                                       ............MORE COMING SOON..........

🛠 Tech Stacks

• Languages: Python
• Libraries: pandas, numpy, matplotlib, seaborn, scikit-learn
• Tools: Google Colab, Jupyter Notebook

---

🔍 Key Learnings

• End-to-end ML project structure
• Data preprocessing (missing values, encoding, scaling)
• Model building (regression, classification)
• Evaluation using metrics and plots
• Use of Pipeline, ColumnTransformer in scikit-learn

---

If you find these projects helpful, consider giving the repo a ⭐️ and sharing it with others.