📊 Data Analytics & Machine Learning Portfolio

Welcome to my Data Analytics Portfolio, where I showcase hands-on projects across EDA, Customer Segmentation, Sentiment Analysis, and Fraud Detection.
These projects demonstrate my skills in Python, Machine Learning, and Data Visualization applied to real-world datasets.

---

🚀 Projects Overview

1️⃣ McDonald’s Nutrition Facts – Exploratory Data Analysis (EDA)

• Goal: Understand nutrition patterns in menu items.
• Techniques: Pandas profiling, histograms, scatter plots.
• Key Skills: Data cleaning, descriptive statistics, visualization.
• 📂 Code | 🖼️ Sample Plots in outputs/

---

2️⃣ Retail Customer Segmentation

• Goal: Group customers by purchasing behavior.
• Techniques: Feature Engineering, K-Means Clustering.
• Visuals: Scatter plots of clusters.
• Outcome: Identified distinct customer groups (e.g., High-value vs Budget shoppers).
• 📂 Code | 🖼️ Cluster Maps in outputs/

---

3️⃣ Data Cleaning Pipeline

• Goal: Ensure data integrity before analysis.
• Steps:
  • Handle missing values
  • Remove duplicates
  • Detect outliers (IQR method)
  • Standardize formats
• 📂 Code

---

4️⃣ Sentiment Analysis (Twitter + Play Store Reviews)

• Goal: Analyze opinions (Positive, Neutral, Negative).
• Pipeline:
  • Text Cleaning
  • TF-IDF Vectorization → convert text to numbers
  • Logistic Regression → classification
  • Evaluation: Accuracy, Confusion Matrix, ROC-AUC
• Extras: WordClouds for Positive/Negative terms
• 📂 Code | 🖼️ WordClouds in outputs/
• 🔗 Kaggle Notebook Demo (replace with your link)

---

5️⃣ Credit Card Fraud Detection

• Goal: Detect fraudulent transactions from highly imbalanced data.
• Techniques:
  • Logistic Regression (baseline)
  • Random Forest (non-linear patterns)
  • Isolation Forest (anomaly detection)
  • SMOTE (oversampling)
• Evaluation: ROC-AUC, Precision, Recall
• 📂 Code | 🖼️ Confusion Matrix in outputs/
• 🔗 Kaggle Notebook Demo (replace with your link)

---

📦 Tech Stack

• Languages: Python
• Libraries:
  • pandas, numpy → data manipulation
  • matplotlib, seaborn → visualization
  • scikit-learn → ML algorithms (Logistic Regression, Random Forest, KMeans, etc.)
  • imblearn (SMOTE) → handle imbalanced data
  • wordcloud → text visualization

---

🖥️ Repository Structure

1️⃣ McDonald’s Nutrition Facts – Exploratory Data Analysis (EDA)

2️⃣ Retail Customer Segmentation

3️⃣ Data Cleaning Pipeline

4️⃣ Sentiment Analysis (Twitter + Play Store Reviews)

5️⃣ Credit Card Fraud Detection