DataScience Portfolio
Recent Projects
Consumer Complaint Classification
This project explores application of different NLP technologies in classification of consumer complaint for various financial products. The data used for this analysis is obtained from CFPB, The Consumer Financial Protection Bureau a U.S. government agency that makes sure banks, lenders, and other financial companies treat consumers fairly.
Repository: https://github.com/kevin-okiah/CustomerComplaintsClassification-NLP
CLassification - Banking Term Deposit
Modeled and build classifiers to predict the odds of a client being subscribed to a banking term deposit based on exploratoryvariables. Use SMOTE , RUS and ROS to impute data for minority class
Repository: https://github.com/kevin-okiah/Banking_Term_Deposit
Sentiment Analysis
Sentiment Analysis used to isolate Positive and Negative IMDB movie Reviews
Repository: https://github.com/kevin-okiah/NaturalLanguageProcessing-NLP/tree/master/SentimentAnalysis
Topic Modeling
LDA used to recommend Gutenberg Books that share the same Topic to my son
Repository: https://github.com/kevin-okiah/NaturalLanguageProcessing-NLP/tree/master/TopicModeling
Death2GridSearch
Utility to select the best performing Classifier with hyperparameters based on Accuracy, Precision, Recall, f1_score and roc_auc_score
Repository:https://github.com/kevin-okiah/Death2GridSearch-Classification
