kabirnagpal/ml-track

This repository is a recommended track, designed to get started with Machine Learning.

SoA-ML-14

Week 1: Intro to Numpy and Pandas

(Anaconda, spyder, jupyter)

Getting Familiar with:

Link to Week 1's Jupyter Notebook

• Numpy
• Pandas
• Matplotlib

Week 2: Basic Data pre-processing:

Link to Week 2's Jupyter Notebook

• One Hot encoding
• Label Encoding
• Normalization
• Dealing with Missing values
• Introduction to Machine learning
• Types of Learning (Supervised, Unsupervised and Reinforcement)
• Application of Machine Learning

Week 3: Regression Algorithms:

Link to Week 3's Jupyter Notebook

• Linear Regression
• Multiple Linear Regression
• Polynomial Regression

Week 4:Classification Algorithms:

Link to Week 4's Jupyter Notebook

• Logistic Regression
• K-Nearest Neighbours
• Support Vector Classifier
• Decision Tree
• Random Forest
• Voting Classifier

Week 5: Bias vs Variance Trade off

Link to Week 5's Jupyter Notebook

• OverFitting
• UnderFitting
• Regularization
• Support Vector Machines

Week 6:Clustering Algorithms:

Link to Week 6's Jupyter Notebook

• K-means Clustering
• Hierarchical Clustering

Week 7: Dimensionality Reduction:

Link to Week 7's Jupyter Notebook

• PCA

• LDA

• Kernel PCA

• Model Selection:

• K-fold Cross Validation

• Parameter Tuning

• Grid Search

Week 8: An introduction to Boosting

Link to Week 8's Jupyter Notebook

• Gradient Boosting
• XGBoost

ml-track's Issues

Create readme

ML Track Conversions

@kabirnagpal Can we interchange week 2 and week 3 as teaching the things which are linked is good. for example, we use a one-hot encode in logistic regression for the classification and label encoding in NLP. so, what you say it is really good I guess people get easy to catch things.

Track Continuation

We want to take the repository forward to 14 Weeks that will roughly account for 100 Days.

• Deep Learning Code shall be included

• NLP processing and code shall be included

• Project links need to be included.