Data Science

Students can analyze the Titanic dataset to explore patterns in survival rates based on factors like gender, age, passenger class, and fare. They'll create visualizations like bar charts, heatmaps, and box plots to summarize findings.

Titanic Dataset (available on Kaggle)

Use customer data (such as annual income, spending score, age) to segment customers into different groups. The project will involve preprocessing the data and applying K-means clustering to identify customer segments.

Mall Customer Dataset (Kaggle)

Students will predict house prices using data on house features like the number of bedrooms, square footage, location, etc. They'll train a linear regression or a decision tree model and evaluate its accuracy.

Ames Housing Dataset or Boston Housing Dataset (Kaggle)

Using a dataset of tweets, students can build a sentiment analysis model to classify tweets as positive, negative, or neutral. They'll apply techniques like tokenization, stopword removal, and vectorization (e.g., TF-IDF) and use models like Naive Bayes or SVM.

Twitter Sentiment Analysis Dataset (Kaggle or Twitter API)

Using a credit card transactions dataset, students can build a model to detect fraudulent transactions. Techniques for handling imbalanced data, such as SMOTE (Synthetic Minority Over-sampling Technique), can be applied.

Credit Card Fraud Detection Dataset (Kaggle)