Projects.

1. ML to Predict Hotel Reviews - Github

Implemented text classification with machine learning algorithms and sentiment analysis with NLP.

• Scraped data from TripAdvisor using Beautiful Soup and Selenium libraries.

• Implemented 2 vectorization methods: Word Count vectorizer and Term Frequency-Inverse Document Frequency.

• Partitioned of the target value into 5, 3 and 2 categories to improve accuracy.

• Built 8 ML classifiers identifying Support Vector Machine as the top performing model with a 87% accuracy score after tuning the parameters using GridSearchCV algorithm.

2. ML to Predict Amazon Reviews Sentiment - Github

Applied machine learning in an NLP classification task to identify positive or negative reviews.

• Scraped Amazon review data to create custom dataset using the Beautiful Soup library.

• Created shaped word clouds that make it easy for audiences to visualize the top words in the reviews.

• Implemented sentiment analysis in my dataset for labelling purposes.

• Built 9 machine learning architectures identifying AdaBoost as the top performing model with a 86% accuracy score.
  

3. ML to Predict Students Performance - Github

Implemented machine learning models to predict student performance.

• Built classification models capable of predicting student performance with 89% accuracy based on students demographic, social and economic characteristics.

• Created 6 machine learning architectures with a grid search, identifying Gradient Boosting as the top performing model.

• Used over-sampling technique (SMOTE) to adjust the class distribution of the target value.