# The Machine Learning Data Science Path

# Why?
I'm currently taking the Impact Theory University Mindset Coaching class, where Tom Bilyeu states success requires formulating a plan.
Therefore, it makes complete sense for me (just like with [Daniel Bourke's Masters' Degree](https://mrdbourke.com/blog/aimastersdegree)) to create my own course outline conveying how I'll take myself from knowing nothing about AI to using it for large-scale change.

# What?
This course should set up the reader with a fundamental understanding of machine/deep learning. Unlike with any one single course out there, this should encompass both the theoretical and practical aspects of AI (difficult, but necessary).

# How?
Although I have a few highly theoretical courses down below, check out [Machine Learning with Phil's math course
breakdown](https://www.youtube.com/watch?v=jL7rudV2vpU).
The following lists below can be shifted around, modified, and altered in any way shape or form for a coherent study plan. Just mix and match the materials with the highest appeal! Take note that multiple sources of similar content (different learning styles/intentions) have been kept in the list, so when one perspective doesn't make sense another will.
Most importantly though, although top-down and bottom-up approaches to learning are equally powerful, **Part 3 CANNOT BE SKIPPED**, as it is the backbone for gaining tangible, real, unique and so irreplaceable experience!

# Stages
## Part 1 - Theoretical Understanding
Here the aim is to gain an overall holistic background of the basic architectures used by the libraries/frameworks extensively latter on.
Here only vectorization libraries are provided.

## Part 2 - Applied Knowledge
This will be a basic overview of how to use implementations of the prior concepts, to solidify one's understanding of how to use AI algorithms.
Here there's an overall focus to demonstrating how machine-learning algorithms can be used in basic situations.

## Part 3 - Realistic Projects
This is the useful part, to apply AI to real-world scenarios, exemplifying its use as a tool for solving complex problems.
Despite learning resources being listed out here, these are easily replaceable (by other self-provided pet projects) when sections 1 and 2 supply adequate foundation knowledge. Although, there is a need to learn about more general data science topics, which may or may not be covered by part 2 (like cleaning/transforming data).

# Topics?
I have a brief list down here, however, I'd like to mention [this article](https://towardsdatascience.com/how-to-learn-data-science-my-path-ba7b9aa94f63) which does a far better job at breaking down the different parts of AI.

## Supervised Learning
1.  Linear Regression
2.  Gradient Descent
3.  Support Vector Machines
4.  Decision Trees
5.  K-Nearest Neighbors
6.  Feature Engineering
7.  Under/Over-Fitting and Generalization
8.  Hyperparameter Tuning
9.  Neural Networks
    - Forward Propagation
    -   Deep Learning
      - Back Propagation
      - Convolutional Networks
      -   Recurrent Networks

## Unsupervised Learning
1.  Logistic Regression
2.  Density Estimation
    - DBSCAN and HDBSCAN
3.  Clustering
    - K-Means
4.  Outlier/Anomaly Detection
5.  Ranking

# Learning Materials
## Courses
### Theoretical Understanding
  - [Machine Learning](https://tinyurl.com/mcd85l5) (Coursera) - in Octave
  - [Foundations of Machine Learning](https://tinyurl.com/y8mz83ox) (Bloomberg)
  - [Deep Learning from the Foundations](https://course.fast.ai/part2) (fast.ai)
  - [Learning From Data](https://tinyurl.com/grdb5bg) (edX)

### Applied Knowledge:
  - [Neural Networks and Deep Learning](https://tinyurl.com/y38jdvut) (Coursera) - shorter course with TensorFlow
  - [Applied Machine Learning in Python](https://tinyurl.com/yy5gw6xd) (Coursers) - longer course teaching variety of libraries
  - [Practical Deep Learning for Coders](https://course.fast.ai/) (fast.ai) - longer course using Fast.AI library (extends PyTorch)

## Books

### Theoretical Understanding:
  - [The Hundred-Page Machine Learning Book](https://amzn.to/2p8meGH) - mostly maths
  - [Learning from Data](https://amzn.to/363fXfW) - has accompanying [MOOC](https://tinyurl.com/grdb5bg) above
  - [Grokking Deep Learning](https://amzn.to/2Pi9CaD) - pure Python

### Applied Knowledge
  - [Hands-On Machine Learning with Scikit-Learn and TensorFlow](https://amzn.to/31IQK7o) - uses Scikit-Learn and TensorFlow

# THANKS FOR READING!
Now that you've heard me ramble, I'd like to thank you for taking the
time to read through my blog (or skipping to the end).