Classify google app store user reviews with large language model (distilbert)
Problem:
- Quickly respond to dissatisfied user reviews, potentially extract application problems to improve user experience.
Solution steps:
- We begin with the data, and find relevant review datasets for our use case.
- Luckily we have Google Play Store reviews on the Kaggle Dataset
- Train (or fine-tune) a machine learning model on the data to analyze reviews for us.
- Use the model to filter reviews quickly and help us respond quickly to dissatisfied reviews.
- Potentially extract user feedback insights to help us improve our app. (that’s out of the scope of this notebook, to be implemented later)
Read more: App User Feedback Analysis (Sentiment Analysis)
Diabetes Classification using SVM, Logistic Regression and KNN
Problem:
- Predict wither a patient has diabetes from diagnostics measures.
Solution steps:
- We begin with the provided dataset link.
- Explore the dataset.
- Preprocessing and feature engineering.
- Train a machine learning model.
- Analyze results and look for improvements.
Read more: Predict the onset of diabetes based on diagnostic measures
Seoul Bike Rental (Data Analysis)
Problem:
- Predict the total count of bike rentals in a given hour (or day).
Solution steps:
- We begin with the provided dataset link.
- Exploratory data analysis.
- Feature selection.
Read more: Seoul Bike Rental Exploratory Data Analysis
Chatbot to handle customer services
Problem:
- Automate customer service to handle the growing customer base.
Solution steps:
- Customers intent pattern data exploration.
- Modeling.
- Build a chatbot module.
- Analysis and further improvements.
Read more: Minimal Sales Chatbot