This Repository contains code used in the 1st Lecture Series for ML Hackathon on Kaggle: Help Autonomous Cars Recognize Street Signs

Here are some Useful Resources

Introduction to Machine Learning

• Introduction to ML slides
• Beginner's Guide to Machine Learning Competitions

Classification:

• Logistic Regression
• Stochastic Gradient Descent
• Intro to feature engineering

Neural networks:

• For classification - https://en.wikipedia.org/wiki/Perceptron
• Intro to tensorflow

General:

• Scikit-learn tutorial
• Use the IPython Jupyter notebook. It helps in managing packages and environments. It's also useful for examining your data at different stages of your algorithm
• A helpful reference: http://scikit-learn.org/stable/_static/ml_map.png

Good Visualizations:

• K-Means
• DBSCAN
• Trees

Some useful documentation:

• Scikit-learn
• Matplotlib
• Scipy
• Pandas

There are various helpful articles provided by Kaggle also. You can check them out too. Plotting the error as you learn is a great way of visualising whether your model is working correctly. Matplotlib.pyplot is very useful for this.

For Fun: Try playing with Orange3 Data Mining Toolkit

---

Advanced section:

Ensembles :

• Basics
• Bagging
• Gradient Boosting
• Adaboost

Hierarchical clustering: https://www.youtube.com/watch?v=GVz6Y8r5AkY&list=PLBv09BD7ez_7qIbBhyQDr-LAKWUeycZtx