This is part of the Competition where you’re challenged to build a machine learning model that predicts which Tweets are about real disasters and which one’s aren’t.
Library Used: Tensorflow, Pandas, NumPy, Matplotlib, scikit-learn, nltk
My Contribution:
- Exploratory Data Analysis
- Preprocessing
- Wordcloud for common words in real and non-real tweets
- Build a DNN model with Bidirectional GRU with public score of 0.71437
This is part of the Competition where you’re challenged to predict the final price of each home WITH With 79 explanatory variables describing (almost) every aspect of residential homes in Ames, Iowa.
Library Used: Pandas, NumPy, Matplotlib, scikit-learn
My Contribution in Notebook-EDA:Ref
- Relationship of 'SalePrice' with numerical and categorical variable
- Heatmap and scatterplot of 'SalePrice' and correlated varaibles
- Outlier identification with Bivariate analysis
- Check for 'normality', 'skewness' and 'homoscedasticity'
- Log Transform data to attain normality
- Convert categorical variable into dummy
My Contribution in Notebook-Modeling
- Data Preprocessing
- Identify and remove outliers
- Log Transform data to attain normality
- Feature Engineering
- Identify and impute missing data
- Transforming some numerical variables that are really categorical
- Label Encoding some categorical variables that may contain information in their ordering set
- Box Cox Transformation of (highly) skewed features
- Modeling
- LASSO, Elastic Net, Kernel Ridge, Gradient Boosting, XGBoost, LightGBM
- Stacking models
- Averaging base models
- Add a meta-model on averaged base models and use the out-of-folds predictions of these base models to train meta-model
- Ensembling StackedRegressor, XGBoost and LightGBM The submission is in Top 6%