This repository showcases a variety of machine learning and data science projects.
This section includes projects focussing on gathering insights using data visualization and dashboarding.
Weather Forecasting This project scrapes data from openweathermap.org site, and generates weather insights for multiple cities in US. Later, these are then displayed in a Tableau interactive dashboard for the user to play around.
This section includes implementations of popular ML algorithms on some kaggle datasets using an OOPS based approach. Additionally, this also includes code to develop machine learning algorithms built from the ground up using Python.
regressors_from_scratch.ipynb: Explore various regression techniques for numerical prediction tasks.
- Linear Regression
- Logistic Regression
kaggle_predicting_housing_prices.ipynb Performing EDA on the housing_prices dataset, developing and testing various models using an OOPS based approach and finding the best performance.
The idea here is to showcase how regressor algorithms can be implemented in a classes and objects OOPS format.
This section includes implementation of several deep learning models:
deep_learning_from_scratch.ipynb Implementing deep learning model on digits dataset from scratch. The idea here is to showcase how forward and backward propagation works, and how to train a model from scratch. kaggle_digit_recognizer.ipynb Implemented tensorflow deep learning mode to improve the val/test accuracy for the digits dataset, and submitted those results in kaggle. Check out my current leaderboard position in the Digit Recognizer.
Note: To look for .py scripts directly, and not notebooks, go to src/ folder.