A web application to perform A/B testing that leverages Bayesian methods as well as Flask, Python, and PyMC3 for statistical analysis. Users can input trial and success data for two variants, and the app models and compares their performance. It provides probabilities showing which variant is likely to perform better, along with visualizations of the results. The app is a useful tool for making data-driven decisions in experiments, product testing, and marketing campaigns

Mar 17, 2024

An interactive subjectivity analysis tool developed in Python, using Natural Language Processing (NLP) to classify sentences into subjective or objective categories while adapting through user feedback. The tool is powered by a Naive Bayes classifier trained on NLTK’s subjectivity dataset, with the ability to adapt and improve based on user feedback. Users can input sentences for classification, and if the prediction is incorrect, they can provide the correct label to refine the model over time. This makes the tool useful for tasks like text analysis, sentiment evaluation, and automated content moderation, while continuously learning and improving from new data.

Mar 12, 2024

A 20 page report on Linear Regression Modeling for a California Housing Dataset. The implementation is in R and uses methods like correlation analysis to explore relationships between variables and multiple linear regression to predict median house values. It also ANOVA for comparing group means, applies Tukey's post-hoc tests for detailed comparisons, and uses stepwise regression to refine and improve the predictive models.

Jun 10, 2022