This project aims to develop a machine learning model to predict whether clients of a Portuguese banking institution will subscribe to a term deposit, utilizing data from direct marketing campaigns involving phone calls.
Techniques and Tools Utilized:
-
Data Analysis: Leveraged pandas for data manipulation and exploration, allowing for in-depth analysis of the dataset.
-
Feature Selection: Employed data science techniques to select the most relevant features, ensuring the model's effectiveness and efficiency.
-
Model Training: Utilized scikit-learn to encode categorical variables and train various machine learning models on the dataset.
Machine Learning Models Used:
-
Multi-Layer Perceptron (MLP): Employed a neural network architecture capable of learning complex patterns in the data.
-
Logistic Regression: Utilized for binary classification, providing insights into the likelihood of term deposit subscription based on predictor variables.
-
Gaussian Classifier: Employed for probabilistic classification, modeling the likelihood of class membership based on Gaussian distributions.
-
Hist Gradient Boosting: Utilized a gradient boosting algorithm capable of handling large datasets and yielding high predictive accuracy.
-
Random Forest Classifier: Employed an ensemble learning technique to aggregate predictions from multiple decision trees, offering robustness and generalization capability.
-
Sequential Layer Neural Network: Implemented a deep learning architecture with sequential layers, enabling the model to capture intricate relationships within the data.
By leveraging these techniques and tools, the project aimed to develop a robust predictive model capable of assisting the banking institution in optimizing its marketing campaigns and improving subscription rates for term deposits.