Pinned

1. PORTFOLIO-Company-bankruptcy-prediction PORTFOLIO-Company-bankruptcy-prediction Public

   The goal of this project is to spot bankrupt companies based on 95 financial book attributes. Methods used: cross validation, shuffle, single feature and incrementally adding feature selection, XGB…

   Jupyter Notebook

2. PORTFOLIO-Credit-score-prediction PORTFOLIO-Credit-score-prediction Public

   The goal of this project is to predict a credit score based on 304 attributes. Methods used: cross validation, feature selection by dropping constant and correlated features and XGBoost Classificat…

   Jupyter Notebook

3. PORTFOLIO-Credit-card-fraud PORTFOLIO-Credit-card-fraud Public

   The goal of this project is to predict if a transaction is fraudulent based on 28 anonymous features and an amount. Methods used: cross validation, XGBoost Classification and SHAP.

   Jupyter Notebook

4. PORTFOLIO-Hungarian-chickenpox-cases-prediction PORTFOLIO-Hungarian-chickenpox-cases-prediction Public

   The goal of the project is to predict chickenpox cases one year ahead based on known history. Methods used: ETS decomposition and SARIMA with statsmodels, LSTM with Keras, MINMAX scaling.

   Jupyter Notebook

5. PORTFOLIO-Air-polution-clustering PORTFOLIO-Air-polution-clustering Public

   The goal of this project is to find areas in India with similar air pollution characteristics. Methods used: agglomerative clustering and DBSCAN with sklearn, MICE, MINMAX scaling.

   Jupyter Notebook

6. PORTFOLIO-Text-topic-classification PORTFOLIO-Text-topic-classification Public

   The goal is to classify the given texts into 4 subjects i.e, maths, physics, chemistry, and biology. Methods used: tokenization, lemmatization, stop words, word clouds, bag of words, neural network…

   Jupyter Notebook 1