1. Students intro
Status Quo?
2. Inner Working of a NN
Neuron
![diagram](static/images/3blue1brown_01.jpg)
Layers
![diagram](static/images/3blue1brown_02.jpg)
Weights
![diagram](static/images/3blue1brown_03.jpg)
Activation & Weights
![diagram](static/images/3blue1brown_04.jpg)
Activation & Weights
![diagram](static/images/3blue1brown_05.jpg)
Sigmoid
![diagram](static/images/3blue1brown_06.jpg)
Bias
![diagram](static/images/3blue1brown_07.jpg)
Learning
![diagram](static/images/3blue1brown_08.jpg)
Matrix operations
![diagram](static/images/3blue1brown_09.jpg)
Matrix operations
![diagram](static/images/3blue1brown_091.jpg)
Neuron is a function
![diagram](static/images/3blue1brown_10.jpg)
NN is a function
![diagram](static/images/3blue1brown_11.jpg)
Sigmoid VS ReLU
![diagram](static/images/3blue1brown_12.jpg)
Cost Function
![diagram](static/images/3blue1brown_14.jpg)
High Cost
![diagram](static/images/3blue1brown_15.jpg)
Low Cost
![diagram](static/images/3blue1brown_16.jpg)
Average Cost
![diagram](static/images/3blue1brown_17.jpg)
NN Function Parameters
![diagram](static/images/3blue1brown_18.jpg)
Cost Function Parameters
![diagram](static/images/3blue1brown_19.jpg)
1D Cost Function
![diagram](static/images/3blue1brown_20.jpg)
2D Cost Function
![diagram](static/images/3blue1brown_21.jpg)
Gradient Descent
Find the shortest path to walk down a hill
One step after another...
Different importance of components
![diagram](static/images/3blue1brown_22.jpg)
Propagating Backward
![diagram](static/images/3blue1brown_23.jpg)
Summary - Looking back at 'our three figures'
Forward Propagation
![diagram](static/images/Presentation1.022.png)
Gradient Descent
![diagram](static/images/Presentation1.023.png)
back Propagation
![diagram](static/images/Presentation1.024.png)
Machine Learning Project Checklist
- 1. Frame the problem and look at the big picture.
- 2. Get the data.
- 3. Explore the data to gain insights.
- 4. Prepare the data to better expose the underlying data patterns to Machine Learning algorithms.
- 5. Explore many different models and short-list the best ones.
- 6. Fine-tune your models and combine them into a great solution.
- 7. Present your solution.
- 8. Launch, monitor, and maintain your system.
Credits
- Photos unsplash.com
- Diagrams 3Blue1Brown
- Checklist Hands-On Machine Learning with Scikit-Learn and TensorFlow