python-ML

My take on various Data Science Competitions/Hackathons. Performing Data Preprocessing, Data Analysis, and creating Machine Learning models to make predictions on real world datasets. Applying Feature Engineering, Deep Learning, Ensemble Learning. Using scikit-learn for ML, Keras with TensorFlow backend for Deep Learning.

List of Machine Learning Competitions/Hackathons:

1. predicting-flight-ticket-prices: MachineHack Predicting Flight Ticket Price Hackathon (Got a score of 94.75 on leaderboards with a simple Random Forest algorithm, leader score 95.5)
   
   
   
   

2. ltfs-data-science-finhack: Analytics Vidya LTFS Data Science FinHack (ML Hackathon). (Used an Ensemble of ANN, LGBM and Random Forest classifiers to get a rank in the top 6% in Vehicle Loan Default Prediction Challenge.)
   
   
   
   

Miscellaneous ML projects:

1. camera-calibration-with-opencv-python: Using OpenCV-Python to perform camera calibration.
   
   

2. kaggle-titanic-dataset: Performing binary classification on Kaggle Titanic dataset.
   Using:

   • Random Forest, Logistic Regression, SVM algorithms.

   • ANN.
     
     
     
     

3. kaggle-dogs-vs-cats-redux: Performing binary image classification on Kaggle Dogs vs Cats Redux dataset. Using Keras to build a CNN and using Data Augmentation.
   
   

• If the notebooks don't render in github, open them using nbviewer.

  • Open the notebook in github.

  • Copy the page URL.

  • Paste the URL in nbviewer.

  • Click on Go. Notebook will open in nbviewer.