Predicting Bankruptcy - Machine Learning Algorithms Gathered financial data of companies across several industries and their different financial ratios from the period of 1980 – 2001 to determine the risk of bankruptcy for a company
- Data from various text files loaded on Big Data Platform Hortonworks Ambari and consolidated using HiveQL
- Conducted initial pre-processing on the data using Pig Script and split the data into train and test set to export as CSV files
- Conducted Exploratory data analysis in Python to identify patterns or trends which could be helpful in cleaning the data and getting insights
- Manipulated the data to handle missing values, outliers and transform the skewness in data to avoid bias in the model
- Applied Machine learning algorithms like Logistic Regression, Neural Networks, Decision Trees and Support Vector Machines to to build the classification model to predict bankruptcy
- Tested Neural Networks to be our final model with Accuracy was 78.21% and Misclassification rate to be 0.2178