Data-Science-Projects:

Techniques:

Feature Selection:

• PCA (Principal Component Analysis)
• AIC (Akiake Information criterion)
• BIC (Bayesian Information criterion)
• LASSO (Least Absolute Shrinkage and Selection Operator)

1. Credit Card Fraud Detection: {Python: Sckit-learn, Tensorflow, R} (Ongoing

• Models:
  1. Random Forest
  2. Gradient Boosting
  3. XGBoost
  4. Deep Neural Nets
  5. Autoencoders
  6. Bayesian Methods

2. Diabetic-Readmission Analysis: {PySpark, R}

• Classification:
  1. GLM {RIDGE/LASSO/ELNET}
  2. Random Forests

3. Crime Prediction: {Python: Sckit-learn}

• Regression:

  1. Linear Regression
  2. Polynomial Regression

• Classification:

  1. Decision Trees
  2. Gaussian Naive Bayes
  3. Support Vector Machines, Linear SVC, POLY, RBF
  4. Random Forests

4. Credit default: {R}:

• Classification:
  1. Logistic Regression (GLM): RIDGE/LASSO
  2. Naive Bayes
  3. Decision Trees
  4. Random Forests

5. Loan Default: {R}

• Classification:
  1. GLM (Generalized Linear Model)

--> Data {source URL} :
1. http://archive.ics.uci.edu/ml/
2. https://www.lendingclub.com/info/download-data.action