30DaysMachine-Learning-Challange-roadmap

This is a tentative roadmap for our 30 days machine learning challange. I will add more information along the way.

Day One:

• Core concepts of Machine Learning

• Machine Learning Process

Day Two:

• Project:
• Classification Walkthrough: Titanic Dataset

Day Three:

• How Machine Learning Models work

Day Four:

• Basic Data Exploration

• Working with Missing Data

a. Examining Missing Data

b. Dropping Missing Data

c. Imputing Data

d. Adding Indicator Columns

Day Five:

• Your first machine-learning model

• Working with Cleaning Data

a. Column Names

b. Replacing Missing Values

Day Six:

• Model Validation

• Data Exploration

a. Data Size

b. Summary Stats

c. Histogram

d. Scatter Plot

e. Joint Plot

f. Pair Grid

g. Box and Violin Plots

h. Comparing Two Ordinal Values

i. Correlation

j. RadViz

k. Parallel Coordinates

Day Seven:

• Underfitting vs Overfitting

• Preprocessing Data

a. Standardize

b. Scale to Range

c. Dummy Variables

d. Label Encoder

e. Frequency Encoding

f. Pulling Categories from Strings

g. Other Categorical Encoding

h. Date Feature Engineering

i. Add col _na Feature

j. Manual Feature Engineering

Day Eight:

• Feature Selection

a. Collinear Columns

b. Lasso Regression

c. Recursive Feature Elimination

e. Mutual Information

f. Principal Component Analysis

g. Feature Importance

Day Nine:

• Dealing with Imbalance Classes

a. Use a Different Metric

b. Tree-based Algorithms and Ensembles

c. Penalize Models

d. Upsampling Minority

e. Generate Minority Data

f. Downsampling Majority

g. Upsampling Then Downsampling

Day Ten:

• Classification Algorithms

Day Eleven:

• Model Selection

Day Twelve:

• Metrics and Classification Evaluation

a. Confusion Matrix

b. Metrics

c. Accuracy

d. Recall

e. Precision

f. F1

g. Classification Report

h. RoC

i. Precision-Recall Curve

j. Cumulative Gains Plot

k. Lift Curve

l. Class Balance

m. Class Prediction Error

n. Discrimination Threshold

Day Thirteen:

• Explaining Classification Model

Day Fourteen:

• Regression Algorithms

Day Fifteen:

• Metrics and Regression Evaluation

Day Sixteen:

• Explaining Regression Model

Day Seventeen:

• Dimensionality Reduction

Day Eighteen:

• Clustering

Day Nineteen:

• Implementing Pipeline

Day Twenty:

• Neural networks

• Artificial neural networks (ANN)

Day Twenty-one:

• Project:

• ANN walkthrough: Predicting Stock Prices

Day Twenty-two:

• Natural Language Processing (NLP)

Day Twenty-three:

• Project:

• NLP walkthrough: Mining Newsgroups Dataset

Day Twenty-four:

• Deep Learning Basics

Day Twenty-five:

• Problems and Solutions

Day Twenty-six:

• Machine Learning best practices

Day Twenty-seven:

• Project:

• Building a Movie Recommendation Engine

Day Twenty-eight:

• Project:

• Recognizing Faces

Day Twenty-nine:

• Project: Predicting Online Ad Click-Through: Tree-based Algorithm

Day Thirty:

• Project: NewsGroups Dataset with Clustering and Topic Modeling

Credit

• Machine Learning 30 Days Challenge