Bank-Marketing-Data-Analysis

Requirements

• Python 2.7
• Numpy >= 1.14.2
• Matplotlib >= 2.2.0
• Pandas >= 0.22.0
• Scikit-Learn >= 0.19.1

Description

1. Bank Marketing dataset is collected from direct marketing campaign of a bank institution from Portuguese.

2. Marketing campaign can be understood as phone calls to the clients to convince them accept to make a term deposit with their bank.

3. After each call, they are being noted as to no - being the client did not make a deposit and yes - being the client on call accepted to make a deposit.

4. The purpose of this project is to predict if the client on call would accept to make a term deposit or not based on the information of the clients.

5. The Bank Marketing Data Set considered for this project is a small portion (10%) of the entire available data set. The data set has about 4119 rows of data with 19 features and 1 column of Class information.

6. The main issues of the dataset are:

   • Preprocessing required to fill unknown values in the dataset

   • Preprocessing required to decide on usage of categorical data along with continuous data

   • The data is class imbalanced (Number of class 1 (yes) is very low when compared to the number of Class 0 (no))

7. Data Analysis Work done for this analysis include:

   • Understanding of features

   • Preprocessing of features

   • K-Nearest Neighbor Classifier

   • Logistic Regression

   • Naïve Bayes

   • SVM

   • Perceptrons

   • Random Forest Classifier

   • Dimensionality Reduction