musyokapatrickmutuku/financial_inclusion_app

The purpose of this app is to predict the number of people in the countries of Kenya, Rwanda, Tanzania and Uganda have or use a bank account. This enables us to understand their level of financial literacy and their ability to access financial services. # Financial Inclusion in Africa ## Project Overview The goal of this project is to build a machine learning model that can predict which individuals in East Africa are most likely to have or use a bank account. This is part of the broader effort to achieve financial inclusion, which means ensuring that individuals and businesses have access to useful and affordable financial products and services. ## Dataset The dataset used in this project contains demographic information and financial service usage data for approximately 33,600 individuals across East Africa. The dataset is provided as part of the Financial Inclusion in Africa challenge hosted on the Zindi platform. ## Approach ### Data Exploration and Preprocessing: Installed the necessary packages Imported the data and performed basic data exploration Generated a pandas profiling report to gain insights into the dataset Handled missing and corrupted values Removed any duplicates Used the boxplot method to identify outliers, and then applied the winsorization method to handle them Encoded categorical features using LabelEncoder ## Model Training and Evaluation: Trained and tested a machine learning classifier based on the preprocessed data, using the Random Forest Classifier algorithm The features used to train the model were: education level, age of respondent, cellphone access, job type, and country The perfromance of the model was as follows: precision recall f1-score support 0 0.94 0.84 0.88 6073 1 0.40 0.67 0.50 985 accuracy 0.81 7058 Using the f1 score metric, 88% of 0's were predicted correctly and 50% of 1's were predicted correctly This concludes that the model was not performing well in predicting 1's which were the folks with bank accounts as compares to predicting zeros, which were folks with no bank account. Thus this suggests an instance of imbalanced data on the target variable ## Streamlit Application: Create …