Internship Tasks Submission Repository ๐
๐ Inside this repository:
- Organized files for each task, with clear documentation and code.
๐ Project Description:
- TASK 2: CREDIT CARD FRAUD DETECTION
Overview - Built a model to identify fraudulent transactions in a dataset of historical credit card transactions to prevent
unauthorized usage & protect both consumers and financial institutions.
Algorithms & Classifier Used - Logistic Regression & Random Forest Classifier
Features- 1) Provided a clear understanding of how individual features contribute to the prediction of fraud.
2) Evaluation of model performance using relevant metrics (accuracy, precision, recall, F1-score).
3) The dataset is split into two parts: Training & Testing to evaluate their performance better.
4) Random Forest is used to capture complex relationships in the data and handling nonlinearities.
- TASK 3: CUSTOMER CHURN PREDICTION
Overview - A model to predict customer churn for a subscription- based service or business. Used historical customer data,
including features like usage behavior and customer demographics.
Algorithms & Classifiers Used - Random Forest and Gradient Boosting to predict churn.
Features- 1) Implementation of Random Forest classifier for churn prediction.
2) Utilization of Gradient Boosting classifier for accurate churn prediction.
3) Evaluation of model performance using relevant metrics (accuracy, precision, recall, F1-score).
- TASK 4: SPAM SMS DETECTION
Overview - A robust solution for combating spam SMS using the power of Multinomial Naive Bayes!
It is a model which distinguishes between legitimate messages and spam messages.
Classifier Used - Naive Bayes and Multinomial Naive Bayes for the prediction.
Features- 1) Preprocessed and cleaned a diverse SMS dataset, transforming text into numerical features for the model.
2) Employed Naive Bayes classifier, to differentiate spam and non-spam messages.
3) Achieved an accuracy of 98% in classifying SMS messages as spam or not spam.