CodSoft Internship – Machine Learning Projects

Welcome to my Machine Learning project repository completed during the CodSoft Internship. This collection demonstrates practical implementation of classification algorithms on real-world datasets, with a focus on fraud detection, customer behavior analysis, and natural language processing.

---

Project Highlights

#	Project	Objective
1️⃣	Credit Card Fraud Detection	Detect fraudulent transactions from anonymized card usage data
2️⃣	Customer Churn Prediction	Predict if a customer is likely to discontinue a service
3️⃣	SMS Spam Detection	Classify text messages as spam or legitimate

---

💳 1. Credit Card Fraud Detection

• Worked on highly imbalanced dataset
• Models: Logistic Regression, Random Forest, XGBoost
• Techniques: SMOTE, Class Weights
• Metrics: ROC-AUC, Precision-Recall, Confusion Matrix 📂 Task2.ipynb

---

📉 2. Customer Churn Prediction

• Telecom-style customer data
• Feature Engineering on demographics and usage patterns
• Models: Decision Tree, Random Forest, SVM
• Evaluation: Accuracy, F1 Score, Classification Report 📂 Task3.ipynb

---

✉️ 3. SMS Spam Detection

• NLP Preprocessing: Lowercasing, Stopword Removal, Lemmatization
• Models: Naive Bayes, Logistic Regression
• Techniques: Bag-of-Words, TF-IDF Vectorization
• Metrics: Accuracy, Precision, Recall 📂 Task4.ipynb

---

Tech Stack

Category	Tools / Libraries
Language	Python
Libraries	pandas, numpy, scikit-learn, matplotlib, seaborn, nltk
Platforms	Google Colab / Jupyter Notebook
ML Techniques	Classification, NLP, Imbalanced Learning

---

Learning Outcomes

• Built end-to-end ML pipelines for diverse problem domains
• Gained hands-on experience with imbalanced data, feature selection, and text classification
• Applied real-world evaluation techniques to validate model performance
• Strengthened my foundation in practical machine learning

---